State of the Nation’s Housing - Joint Center for Housing · PDF fileThe opinions expressed in The State of the Nation’s Housing 2016 do not necessarily represent the views of Harvard

J O I N T C E N T E R FO R H O U S I N G S T U D I E S O F H A R VA R D U N I V E R S I T Y

THE STATE OF THE NATION’S HOUSING

2016

HARVARD GRADUATE SCHOOL OF DESIGN

HARVARD KENNEDY SCHOOL

Principal funding for this report was provided by the Ford Foundation and the Policy Advisory Board of the Joint Center for Housing Studies. Additional support was provided by:

Federal Home Loan Banks

Housing Assistance Council

MBA’s Research Institute for Housing America

National Association of Home Builders

National Association of Housing and Redevelopment Officials (NAHRO)

National Association of REALTORS®

National Council of State Housing Agencies

National Housing Conference

National Housing Endowment

National Low Income Housing Coalition

National Multifamily Housing Council

© 2016 by the President and Fellows of Harvard College.

The opinions expressed in The State of the Nation’s Housing 2016 do not necessarily represent the views of Harvard University, the Policy Advisory Board of the Joint Center for Housing Studies, the Ford Foundation, or the other sponsoring organizations.

JOINT CENTER FOR HOUSING STUDIES OF HARVARD UNIVERSITY

CONTENTS

Executive Summary 1

Housing Markets 7

Demographic Drivers 13

Homeownership 19

Rental Housing 25

Housing Challenges 31

Appendix Tables 37

EXECUTIVE SUMMARYEXECUTIVE SUMMARY

1JOINT CENTER FOR HOUSING STUDIES OF HARVARD UNIVERSITY

HOUSING RECOVERY SCORECARDBy many measures, the US housing market has recovered sub-stantially from the crash. According to CoreLogic estimates, nominal home prices were back within 6 percent of their previ-ous peak in early 2016, although still down nearly 20 percent in real terms. The uptick in nominal prices helped to reduce the number of homeowners underwater on their mortgages from 12.1 million at the end of 2011 to 4.3 million at the end of 2015. Delinquency rates also receded, with the share of loans entering foreclosure down sharply as well.

But at 1.1 million units, new home construction was still run-ning near historic lows last year. A key factor holding back housing starts is the sustained falloff in household growth. Given the size and age of the adult population and under nor-mal economic conditions, roughly 1.2 million net new households would have formed on average each year in 2007–2013. But the actual increase was just half that number as the weak economy made it difficult for young adults to live on their own and for immigrants to settle in the United States. In 2015, how-ever, with the economy nearing full employment and incomes beginning to climb, household growth returned to its expect-ed pace and new home construction was up by a healthy 11 percent (Figure 1).

Now in its seventh year, the US economic recovery shows signs of flagging in the face of a strong dollar, a weakening global economy, and low energy prices. But as household growth continues to gain momentum, the housing sector should be an engine of growth. Factoring in the need to replace older units and meet demand for vacation homes and other uses, housing construction should average at least 1.6 million units a year over the next decade. This level of activity would provide an important spur to the economy. Indeed, residential fixed invest-ment (including homeowner improvements) has accounted for just 2.8 percent of annual GDP so far this decade, significantly less than the 4.3 percent share averaged in the 1980s and 1990s, leaving plenty of room for growth.

With household growth finally

picking up, housing should help

boost the economy. Although

homeownership rates are still

falling, the bottom may be in

sight as the lingering effects

of the housing crash continue

to dissipate. Meanwhile, rental

demand is driving the housing

recovery, and tight markets

have added to already pressing

affordability challenges. Local

governments are working to

develop new revenue sources to

expand the affordable housing

supply, but without greater

federal assistance, these efforts

will fall far short of need.

THE STATE OF THE NATION’S HOUSING 20162

HOMEOWNERSHIP DOWN BUT NOT OUTThe US homeownership rate has tumbled to its lowest level in nearly a half-century. The decade-long declines are especially large among the age groups in the prime first-time homebuying years (Figure 2). The falloff in homeownership has more than offset earlier gains, leaving age-specific rates for all but the old-est households significantly lower than in 1995.

But a closer look at the forces driving this trend suggests that the weakness in homeownership should moderate over the next few years. A critical but often overlooked factor is the role of foreclosures in depleting the ranks of homeowners. Indeed, CoreLogic estimates that more than 9.4 million homes (the majority owner-occupied) were forfeited through foreclosures, short sales, and deeds-in-lieu of foreclosure from the start of the housing crash in 2007 through 2015.

Although completed forfeitures have slowed considerably, they remain elevated at 670,000 or about twice the annual average before the downturn. In addition, Mortgage Bankers Association (MBA) data indicate that the share of loans that are seriously delinquent (90 or more days past due or in foreclosure) has also fallen sharply, but is still nearly double the average in the first half of the 2000s. Given the current rate of recovery, foreclo-sures are likely to keep downward pressure on homeownership rates for the next two years.

Just as exits from homeownership have been high, transitions to owning have been low. Tight mortgage credit is one explanation, with essentially no home purchase loans made to applicants with subprime credit scores (below 620) since 2010 and a sharp retreat in lending to applicants with scores of 620–660 compared with the early 2000s. And given that the homeownership rate tends to move in tandem with incomes, the 18 percent drop in real incomes among 25–34 year olds and the 9 percent decline among 35–44 year olds between 2000 and 2014 no doubt played a part as well.

● Household Growth ● Housing Starts

2.25

2.00

1.75

1.50

1.25

1.00

0.75

0.50

0.25

0

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Housing Construction Has Picked Up in Line with Household GrowthMillions

FIGURE 1

Source: US Census Bureau, Housing Vacancy Surveys and New Residential Construction data.

Source: US Census Bureau, Housing Vacancy Surveys.

● 1995 ● 2005 ● 2015

90

80

70

60

50

40

30

20

10

025–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–74 75 and Over

Homeownership Rates for Most Age Groups Have Fallen Well Below Pre-Boom LevelsHomeownership Rate (Percent)

FIGURE 2

Age GroupAge Group


The good news for the owner-occupied housing market is that these constraints should ease as the mortgage market continues to wrestle with the fallout from the housing crash and adapts to a new regulatory environment. There are already indications from the Federal Reserve’s Senior Loan Officer Opinion Survey that credit standards may be loosening, particularly for loans backed by the government-sponsored enterprises (GSEs). The upturn in real income growth among younger households should also help.

Other structural shifts, however, could also have an impact on homeownership rates—in particular, the rising tide of student loan debt. The share of adults aged 20–39 with student loan debt soared from 22 percent in 2001 to 39 percent in 2013, while the average amount that borrowers owed jumped from $17,000 to $30,000 in real terms. Although student loan payments should not limit the homeownership options of most households, this may not be true for the nearly one-fifth of indebted young renters whose pay-ments exceed 14 percent of monthly income, a level the Consumer Financial Protection Bureau considers highly burdensome.

Several long-term demographic forces are also at work. Ages at first marriage and the start of childbearing have been on the rise for some time, implying delays in first-time homebuying. The growing minority share of the population also has a dampening effect, given minorities’ much lower homeownership rates. At the same time, though, the aging of the baby-boom generation (born 1946–1964) is increasing the share of households over 50, the ages when homeowning is most common. On net, these countervailing trends are unlikely to move the homeownership rate much.

The bigger question is whether the housing crash diminished the general appeal of homeownership. The available evidence sug-gests that it has not. For example, a 2015 Demand Institute survey of more than 5,000 households found that 89 percent of respon-dents under the age of 30 owned a home, would buy a home on their next move, or would buy a home in the future. The shares of respondents with similar responses exceeded 80 percent in all other age groups as well. In addition, 63 percent of all respondents to the April 2016 Fannie Mae National Housing Survey also stated that they would buy homes on their next move.

In short, the near-term direction of the US homeownership rate will depend more on whether households can finance their pur-chases than whether they have the desire to own. Over the next few years, homeownership will continue to face the headwinds created by a backlog of homes in foreclosure, tight credit, weak income growth, and impaired credit histories. But as these pres-sures ease, there is every reason to expect homeownership rates to show some increase.

RENTAL MARKET STRENGTH The rental market continues to drive the housing recovery, with over 36 percent of US households opting to rent in 2015—the largest share since the late 1960s. Indeed, the number of renters increased by 9 million over the past decade, the largest 10-year gain on record. Rental demand has risen across all age groups, income levels, and household types, with large increases among older renters and families with children.

Notes: Rents are from the CPI rent index for primary residence. Changes in vacancy rates are based on a four-quarter trailing average.Source: JCHS tabulations of US Bureau of Labor Statistics, Consumer Price Indexes and Census Bureau, Housing Vacancy Surveys.

● Rent Index (Left scale) ● Vacancy Rate (Right scale)

5

4

3

2

1

0

-1

-2

-3

-4

-5

1.25

1.00

0.75

0.50

0.25

0.00

-0.25

-0.50

-0.75

-1.00

-1.25

20042000 20022001 2006 2008 201120102003 2005 2007 2009 2013 2015 20162012 2014

Vacancy Rates Have Fallen for Five Full Years, Pushing Up Rents Year-over-Year Change (Percent)

FIGURE 3

Year-over-Year Change (Percentage points)


Although conversion of formerly owner-occupied single-family homes to rentals met much of the initial surge in demand, construction of multifamily units is now taking on a growing share. But even with these additions to sup-ply, rental vacancy rates have fallen steadily since 2010, dropping to just 7.1 percent by the end of 2015. Rents have climbed in response, with the Consumer Price Index for rent on primary residences up 3.6 percent in nominal terms last year (Figure 3). When adjusted for inflation, it has been three decades since either of these measures registered such tightness in the rental market.

A growing supply of new housing in the pipeline may help ease these conditions, although most new units are intended for the upper end of the market. The median asking rent on new apartments was $1,381 per month in 2015, well out of reach for the typical renter earning $35,000 a year. High rents reflect several market conditions, including a limited supply of land zoned for multifamily use and a complex approval process that adds to development costs. Perhaps most important, however, is growing demand from higher-income households.

Concerns are increasing that multifamily property valuations in some markets may be overinflated. The strong financial performance of rental properties and the relatively low yields from competing investments have driven up demand, pushing the Moody’s/RCA price index for investment-grade properties 39 percent above the previous high. Capitalization rates are now below levels at the height of the housing boom. Valuations are particularly high in the New York metro area, where property values were 93 percent above their previous peak in the fourth quarter of 2015, and in San Francisco, where they were 85 per-cent above peak.

COST-BURDENED RENTERS AT HISTORIC HIGHSThe divergence between the rental and owner-occupied mar-kets is evident in the number of cost-burdened households in each segment. On the owner side, the number of households facing cost burdens (paying more than 30 percent of income for housing) has fallen steadily as high foreclosure rates have pushed out many financially strained owners, low interest rates have allowed remaining owners to reduce their housing costs, and fewer young households have moved into homeownership. As of 2014, the number of cost-burdened owners stood at 18.5 million, down 4.4 million since 2008.

The decline has occurred across all age groups, but especially among younger homeowners. Homeowners age 75 and over, however, are among the most cost-burdened groups, with their share at 29 percent compared with 24 percent for households under age 45. With the aging baby boomers swelling the ranks of older homeowners and larger shares of households carrying mortgage debt into retirement, the problem of housing cost burdens among the elderly is likely to grow.

On the renter side, the number of cost-burdened households rose by 3.6 million from 2008 to 2014, to 21.3 million. Even more troubling, the number with severe burdens (paying more than 50 percent of income for housing) jumped by 2.1 million to a record 11.4 million. The severely burdened share among the nation’s 9.6 million lowest-income renters (earning less than $15,000) is particularly high at 72 percent. In all but a small share of markets, at least half of lowest-income renters have severe housing cost burdens (Figure 4). While nearly universal among lowest-income households, cost burdens are rapidly spreading among moderate-income households as well, espe-cially in higher-cost coastal markets.

HOUSING ASSISTANCE STRUGGLING TO KEEP UP Most federal housing assistance is targeted to very low-income households (earning 50 percent or less of area median). Some 18.5 million renters met this criterion at last count in 2013, up 2.6 million since 2007. Meanwhile, the number of renters receiv-ing some form of assistance from the US Department of Housing

Notes: Severely cost-burdened households pay more than 50% of income for housing. Data are for core based statistical areas (CBSAs).

Source: JCHS tabulations US Census Bureau, 2014 American Community Survey 1-Year Estimates.

● 25–49 ● 50–59 ● 60–69 ● 70–79 ● 80–99

In Most of the Country, a Large Majority of Lowest-Income Renters Are Severely Cost Burdened

FIGURE 4

Share of Renters with Incomes Under $15,000 with Severe Burdens (Percent)


and Urban Development (HUD) actually fell by 159,000 from 2007 to 2013, with a loss of project-based units more than offsetting an increase in housing vouchers.

Only one in four income-eligible renters receives assistance of any kind, leaving millions to try to find housing they can afford in the private market. But units affordable to lowest-income house-holds are often already occupied by higher-income households. Indeed, the National Low Income Housing Coalition estimates that only 57 units were affordable and available for every 100 very low-income renters in 2014. The shortfall for extremely low-income households (earning 30 percent or less of area median) is even more acute, with just 31 housing units affordable and avail-able for every 100 of these renters.

The lack of a strong federal response to the affordability crisis has put new pressure on state and local governments to act. A number of cities have now developed plans to expand afford-able housing options for a broad spectrum of renters, from those facing homelessness up to middle-income households. In addition to federal funds, these plans draw on a range of resources that include linkage fees on new commercial devel-opment, tax-increment financing, taxes on real estate transac-tions, and the use of publicly owned land.

Another increasingly common approach is the adoption or expansion of inclusionary zoning ordinances, either mandating

that a share of new units have below-market-rate rents or offer-ing the opportunity for higher development densities in exchange for affordable set-asides. But cities can only go so far on their own. Recent estimates from the Lincoln Institute of Land Policy show that inclusionary housing programs produced just 129,000–150,000 affordable units nationwide from the 1970s through 2010, making a strong federal support system still essential.

CONSEQUENCES OF HIGH-COST HOUSING The lack of affordable housing options forces cost-burdened renters to sacrifice other basic needs, settle for inadequate liv-ing conditions, and/or face housing instability—all with serious immediate and long-term consequences. The most significant cutback low-income households make is on basic sustenance. Compared with otherwise similar households able to find hous-ing they can afford, severely burdened households in the bot-tom expenditure quartile spend $150 (41 percent) less on food each month. They also spend substantially less on healthcare and put aside less for retirement.

Another tradeoff is between housing that is affordable and housing that is adequate. In 2013, 10 percent of low-income renters lived in units that lacked complete plumbing or kitchen facilities, experienced frequent breakdowns in major systems, or had other physical defects. Housing quality issues are preva-lent in non-metro areas and tribal lands, where the housing stock is more likely to be substandard.

Housing cost burdens also expose renters to the risk of eviction, with all its damaging impacts on household finances, employ-ment prospects, and school performance. In 2013, 2.1 million low-income renters reported that they had missed a rent pay-ment in the previous three months, and a similar number stated they believed they were likely to face eviction in the next two months (Figure 5). Meanwhile, about 710,000 renters had been threatened with eviction in the previous three months, with nearly eight out of ten of these threats associated with a failure to pay rent or other lease violations.

The costs of housing instability are high not just for individual households, but also for the government programs ultimately needed to support homeless families. But recent research has found that providing assistance for permanent housing for homeless families can help reduce domestic violence and sub-stance abuse, keep families together, and limit the number of school moves for children. Importantly, the provision of perma-nent housing is more cost-effective than helping these families through the shelter system.

THE GROWING CONCENTRATION OF POVERTY One enduring legacy of the Great Recession is the further concentration of poverty. In 2000, 6.5 million Americans lived

Notes: Extremely/very/low-income households earn up to 30%/31–50%/51–80% of area medians. Rent payments were missed within the previous three months. Source: JCHS tabulations of US Department of Housing and Urban Development (HUD), 2013 American Housing Survey.

Income Group ● Extremely Low ● Very Low ● Low

Missed Recent Rent Payment(s)

Felt Under Threat of Eviction

Threatened with Eviction

1.2

1.0

0.8

0.6

0.4

0.2

0

Extremely Low-Income Renters Are Especially at Risk of EvictionHouseholds Reporting Housing Insecurity in 2013 (Millions)

FIGURE 5

in neighborhoods with poverty rates of at least 40 percent. In 2014, the population in these areas had more than doubled to 13.7 million, with substantial increases across all racial and ethnic groups (Figure 6). Even so, income disparities as well as still-high levels of racial segregation have consigned 25 percent of poor blacks and 18 percent of poor Hispanics to high-poverty communities, compared with only 6 percent of poor whites.

The consequences of this isolation are profound, particularly for children. Harvard University’s Equality of Opportunity Project has shown that neighborhood conditions deeply affect a child’s success in life, as well as the life expectancy of adults. Indeed, each year spent living in a low-poverty community increases the chances that a child will attend college and have higher lifetime earnings. In addition to these economic benefits, more inclusive communities benefit residents through safer and healthier envi-ronments, including improved air and water quality.

In recognition of these impacts, HUD strengthened its fair housing regulations by issuing a final rule on Affirmatively

Furthering Fair Housing in 2015. The rule requires state and local governments that receive HUD funds, along with all public housing agencies, to identify patterns of segregation in assisted housing and set priorities for addressing disparities. At the same time, a recent Supreme Court ruling on disparate impacts may help to increase the location of new Low Income Housing Tax Credit (LIHTC) units in higher-opportunity communities.

But efforts to broaden the availability of affordable housing in better communities must be viewed within the context of growing segregation by income in the private housing market. According to the Lincoln Institute of Land Policy, more than 500 local jurisdictions have implemented inclusionary housing poli-cies, but the scale and intensity of these programs vary widely and have yet to reach the scale of federal programs.

THE OUTLOOKWhile the rental market continues to expand at a robust pace, the owner-occupied market is still in the process of recovery. Home prices have rebounded sharply in several markets, but they also remain depressed in other areas, leaving millions of owners still underwater on their mortgages. Foreclosures have fallen steadily, but the share of owners seriously delinquent on their loans remains roughly twice what it was before the down-turn. Household credit and balance sheets will take more time to fully heal. Growth in homeowner demand is therefore likely to remain moderate over the next few years as these headwinds finally abate.

But with household growth projected to average over 1.3 million annually over the coming decade, housing construction should continue to climb and help keep the overall economy on solid footing. In addition, the homeownership rate should at least stabilize in the next few years as foreclosures ebb, mortgage credit conditions improve, and household incomes rise.

As it is, however, the need for more affordable rental housing is urgent. The record number of renters paying more than half their incomes for housing underscores the growing gap between market-rate costs and the rents that millions of households can afford. Governments at all levels must redouble their efforts to expand the affordable supply. And with growing recogni-tion that children’s lifelong achievement rests on stable, safe, and healthy living conditions, policymakers must also ensure better access of minority and low-income households to higher-opportunity communities.


Notes: White, black, and other households are non-Hispanic. Hispanic households may be of any race. Other includes Native Hawaiian and other Pacific Islanders, American Indians, Native Alaskans, and people of two or more races.Source: JCHS tabulations of US Census Bureau, 2000 Decennial Census and 2010–2014 American Community Survey 5-Year Estimates.

● 2000 ● 2010–2014

White OtherBlack Hispanic

6

5

4

3

2

1

0

Race/Ethnicity

The Number of People Living in Concentrated Poverty Has More than Doubled Since 2000Population Living in Census Tracts with Poverty Rates of 40 Percent or More (Millions)

FIGURE 6

HOUSING MARKETS


After a mixed year in 2014, the

national housing recovery gained

traction in 2015. Residential

construction continued to climb

as single-family starts revived.

Sales of both new and existing

homes also increased, and likely

would have been even stronger if

inventories were not so low. The

widespread rise in home prices

benefited millions of underwater

homeowners and spurred

renewed investment in homes

and rental properties. With this

rebound, the housing sector has

increased its contribution to the

economy, with more room to grow.

CONSTRUCTION GAINING MOMENTUMHomebuilding remained on the upswing in 2015, with total housing starts climbing 10.8 percent to 1.1 million units (Figure 7). Single-family starts reached the 715,000 mark while completions hit 647,900 units, their highest level since 2008. Even so, the single-family sector is still struggling to recover after a decade of weakness, with only 750,000 units completed annually on average between 2006 and 2015—the lowest number in any 10-year period since 1968. But single-family construction is set to expand thanks to an 8.7 percent increase in permits, to 696,000 units. In fact, single-family permitting accelerated in 2016, averaging 730,000 units at a seasonally adjusted annual rate in the first four months of the year.

On the multifamily side, all key construction measures rose by double digits. Growth in multifamily starts topped 10 percent for the fifth consecutive year in 2015, reaching a 27-year high of 397,300 units. With single-family construction still recovering, 2015 was the fourth consecutive year that multifamily units accounted for more than 30 percent of housing starts, compared with 20 percent on average between 1990 and 2010. Signaling further expansion, multifamily permits rose 18.2 percent last year, to 486,600 units.

Overall construction activity expanded nationwide, with per-mitting up in 70 of the 100 largest metro areas. Just over a third of these metros issued more permits in 2015 than their annual averages in the 1990s, and 20 issued more than their annual averages in the early 2000s. New York was the stand-out, with permits (primarily for multifamily units) soaring 80 percent in 2015, due in part to the impending expiration of a tax abatement program. But permitting in Dallas, Los Angeles, Miami, San Diego, and San Francisco also increased more than 25 percent last year. In contrast, several of the markets that had rebounded quickly after the recession saw permitting slow, including Washington, DC (down 7 percent), Houston (down 11 percent), San Antonio (down 24 percent), and San Jose (down 42 percent).


CHARACTERISTICS OF THE NEW STOCKSingle-family homes are getting bigger, with the median size in 2015 a record-setting 2,467 square feet. Indeed, only 135,000 single-family homes completed in 2014, or about a fifth, were under 1,800 square feet—the lowest number and the smallest share of units this size going back to 1999 (Figure 8). The majority (58 percent) of single-family construction between 2000 and 2014 occurred in low-density urban areas, with another 25 percent built in mid-density urban neighborhoods, 6 percent in high-density urban neighborhoods, and 12 percent built in rural areas.

Meanwhile, the median size of multifamily units fell from nearly 1,200 square feet at the 2007 peak to 1,074 square feet in 2015, reflecting the shift in the focus of development from the owner to the rental market. Many new multifamily units are in large structures, with nearly half of the units completed in 2014 in buildings with 50 or more apartments. In addition, a majority of newly constructed units were located in dense urban areas. Indeed, about 36 percent of all new multifamily units added between 2000 and 2014 were in high-density neighborhoods, and another 30 percent each in medium- and low-density sec-tions of metro areas. Even so, growth in the multifamily housing stock during this period was even more rapid in rural areas (up 24 percent) than in urban areas (up 19 percent).

THE DEVELOPMENT LANDSCAPEThe gradual recovery in single-family construction largely reflects weak demand in the face of sluggish income growth and tight mortgage credit. But constraints on land, labor, and lend-ing may also play a role. Metrostudy data show that the supply of construction-ready land (vacant developed lots) in 50 metro areas shrank by 30 percent from 2008 to 2013, before settling just above levels posted in the early 2000s.

Land supply is firming across metro areas, including those with significant excesses during the housing bubble. In major Florida metros, for example, the average months supply of vacant developed lots soared after 2006, dropped precipitously after 2009, and stabilized in 2015 at 34 months—within the 24–36 month range considered normal. While experiencing milder cycles, major metros in California and Texas had only about a 20-month supply of vacant developed land in 2015, raising the possibility of future constraints on building activity. Land avail-ability in these large states, among others, thus bears watching.

Labor shortages could also be a damper on construction activity. More than 2 million workers left the industry between 2007 and 2013, reducing the construction workforce to 80 percent of its 2007 peak. According to a Census Bureau analysis, only 40 percent of those who lost their jobs between 2006 and 2009 had returned to their previous positions or to other jobs in the industry. Of the remaining displaced workers, more than half found work outside construction and the rest did not return to the formal labor force.

Source: JCHS tabulations of US Census Bureau, New Residential Construction data.

Square Footage ● Under 1,800 ● 1,800–2,999 ● 3,000 and Over

800

700

600

500

400

300

200

100

0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 20141999

Construction of Smaller Single-Family Homes Has Yet to Rebound New Single-Family Homes Completed (Thousands)

FIGURE 8

Key Housing Market Indicators Point to Strengthening in 2015

FIGURE 7

2014 2015

Percent

Change

2014–15

Residential Construction (Thousands of units)

Total Starts 1,003 1,112 10.8

Single-Family 648 715 10.3

Multifamily 355 397 11.8

Total Completions 884 968 9.5

Single-Family 620 647 4.5

Multifamily 264 320 21.2

Home Sales

New (Thousands) 437 501 14.6

Existing (Millions) 4.9 5.3 6.3

Median Sales Price (Thousands of dollars)

New 283.1 296.4 4.7

Existing 208.5 222.4 6.6

Construction Spending (Billions of dollars)

Residential Fixed Investment 550.6 600.1 9.0

Homeowner Improvements 134.8 147.8 9.6

Notes: Components may not add to total due to rounding. Dollar values are adjusted for inflation by the CPI-U for All Items.Sources: US Census Bureau, New Residential Construction and New Residential Sales data; National Association of Realtors®, Existing Home Sales; Bureau of Economic Analysis, National Income and Product Accounts.


This contraction left the construction workforce significantly older. The share of trades workers age 55 and over rose from 10 percent in 2007 to 16 percent in 2013, while the share under the age of 35 fell from 43 percent to 35 percent. This pattern reflects not only the aging of workers that held onto their jobs through the recession, but also a falloff in hiring of younger workers. Indeed, only 13 percent of newly hired construction workers in 2013 were under age 25, down from 18 percent before 2006. Without younger workers to bolster the ranks as older workers move toward retirement, labor shortfalls may emerge. As it is, a 2015 National Association of Home Builders (NAHB) survey found that a majority of construction firms were already report-ing labor shortages in many trades.

To help rebuild its diminished workforce, the construction industry may have to reevaluate the composition of its labor pool. Given that more than a quarter of workers in the trades in 2013 were foreign-born, unpredictable changes in immigra-tion could have an outsized impact on the availability of skilled labor. At the same time, women make up less than 3 percent of trades workers and thus represent a largely untapped resource for the industry.

Meanwhile, development financing is recovering from a sharp drop-off during the recession. According to NAHB, residential construction loan volumes were up 4.5 percent in the fourth quarter of 2015, marking 11 consecutive quarters of increases. While growing nearly 19 percent for the year as a whole, the residential construction loan stock remained 70 percent below the 2008 peak. Other types of acquisition, development, and construction loans have recovered more fully and now stand 51

percent below peak. Credit may be tightening, however. NAHB financing surveys indicate that credit easing slowed at the end of 2015, while the Federal Reserve Board’s Senior Loan Officer Opinion Survey on Bank Lending Practices reported some tight-ening of lending criteria for construction and development loans in late 2015 and early 2016.

STRENGTHENING HOME SALESAfter a slow year in 2014, sales of new single-family homes rose by a robust 14.6 percent in 2015. At just 501,000 units, however, sales remained well below averages in previous decades (Figure 9). Sales of existing homes also rebounded from their 2014 decline, up 6.3 percent to just under 5.3 mil-lion units. Although the rollout of new mortgage disclosure regulations in October led to a temporary dip, existing home sales closed 2015 on a strong note.

Encouragingly, sales of non-distressed properties are driving growth. CoreLogic reports that real-estate owned (REO) and short sales fell by 10–11 percent in 2015, while non-distressed resales rose by 7.6 percent. With this shift, distressed sales accounted for just over 12 percent of existing home sales in 2015, down from 14 percent a year earlier and 28 percent at the peak in 2009–2011. In addition, cash sales (often to inves-tors) accounted for about a third of home purchases last year, the lowest share since 2008 but still well above the pre-crisis average of 25 percent. Meanwhile, the National Association of Realtors (NAR) reports that the first-time buyer share of home sales slipped for the third straight year to 32 percent, its lowest level since 1987.

Sources: JCHS tabulations of NAR, Existing Home Sales and US Census Bureau, New Residential Sales data.

● Existing Homes (Left scale) ● New Homes (Right scale)

8

7

6

5

4

3

2

1

0

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0

200419961995 200019991997 1998 2002200119911990 1993 19941992 2006 2008 201120102003 2005 2007 2009 2013 20152012 2014

New Home Sales Are Still at Historic Lows Units Sold (Millions)

FIGURE 9


LOW INVENTORIES AND VACANCY RATES The stock of existing homes for sale declined 1.9 percent last year, to 2.1 million units. Supply stood at 4.8 months, making 2015 the fourth consecutive year that inventories held below the 6.0-month level, the conventional measure of a balanced market (Figure 10). In contrast, the inventory of new homes for sale climbed 8.2 percent, ending the year at 217,000 units. But even with three years of significant growth, the supply of new homes slipped to just 5.2 months.

One factor keeping first-time homebuyers on the sidelines is that the stock of affordable homes for sale is extremely limited. According to Zillow, inventories of metro area homes in both the bottom- and middle-value tiers shrank by more than 38 per-cent in 2010–2015, while those in the top tier fell by 31 percent. In 2014–2015 alone, bottom- and middle-tier inventories were each down 9 percent, while top-tier inventories declined by 3 percent. As a result, less than 20 percent of existing homes for sale in some of the nation’s largest metros—including Dallas, Denver, Nashville, Phoenix, and Raleigh—were in the most affordable value tier for their areas.

The number of vacant units for sale also declined 1.7 percent from 2014, to 1.4 million. Vacant units for rent were down 3.7 percent, to 3.3 million, adding to rental market tight-ness. The number of vacant units held off market, however, remained elevated at 7.2 million, or 55 percent of all year-round vacant homes. Over half of these vacant units are clas-sified as “other.” According to the Housing Vacancy Survey, about 7 percent of these “other” units were in foreclosure while another 5 percent were involved in other legal actions.

Fully 25 percent were held off market for personal or family reasons, while 16 percent were in need of repair and about 6 percent were abandoned, condemned, or to be demolished. Just under one in ten were undergoing repairs. The large stock of vacant off-market housing may therefore reflect the overhang of distressed properties, as well as the reluctance of some owners to either invest in or sell their units as the market continues to recover.

Overall vacancy rates for both for-sale and for-rent homes are low. After hovering between 2.4 percent and 2.9 percent in 2006–2011, the vacancy rate for owner units fell back in line with longer-term averages in 2015, standing at 1.8 percent for the year. In contrast, the rental vacancy rate plunged from the double-digits in the mid-2000s to a 30-year low of just 7.1 per-cent last year.

PROPERTY PRICES ON THE RISEHome prices continued to climb last year. NAR reports that the median price of existing homes rose for the fourth straight year, to $222,400—a 6.6 percent increase in real terms from 2014 and the highest level since 2007. Meanwhile, nominal home prices reached a new peak in 2015. The CoreLogic, S&P/Case-Shiller, and OFHEO indexes, which are less affected than the median price by the mix of homes sold, show that prices of repeat sales were up 5.3–5.7 percent through the end of last year.

The median new home price increased 4.7 percent in real terms to $296,400 in 2015, topping the 2005 peak. In nominal terms, new home prices were up for the sixth consecutive year, while the Census Bureau’s constant quality index hit a new high. With the number of distressed sales continuing to fall, the gap between new and existing home prices narrowed somewhat from 37 percent on average in 2011–2014 to 33 percent in 2015, but remains relatively wide. By comparison, the average price disparity in the 1990s was just 18 percent.

Home prices in all 20 metro areas tracked by S&P/Case-Shiller were up last year, with increases ranging from under 3 percent in Chicago, Cleveland, and Washington, DC, to about 10 percent in Portland, San Francisco, and Seattle. Prices are now rising across markets that experienced widely different cycles (Figure 11). In Los Angeles and Las Vegas, where significant house price inflation in the mid-2000s was followed by sharp declines, pric-es are again rising rapidly. In the case of Denver, home prices rose only moderately in the first decade of the 2000s but have now climbed to a new high. And in Detroit, where price appre-ciation was modest but the ensuing drop was large, home prices reached an eight-year high last year.

In some metro areas, home values are rising in every tier. In Los Angeles, for example, average nominal increases for all three tiers of zip codes (ranked by median home value in January 2000) topped 150 percent in 2000–2015. Appreciation in Denver

● For-Sale Inventory (Left scale) ● Months Supply (Right scale)

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0

12.0

10.5

9.0

7.5

6.0

4.5

3.0

1.5

0

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

The Number of Existing Homes For Sale Dipped Again in 2015Millions of Units

FIGURE 10

Months

Source: JCHS tabulations of NAR, Existing Home Sales.


also averaged at least 70 percent in every tier, with values in 93 percent of zip codes at new peaks in 2015. The recovery in Las Vegas is more mixed, with values up an average of 53 percent in the top tier, 47 percent in the middle tier, and 31 percent in the bottom tier. And in Detroit, top-tier values rose 15 percent on average over this period, while middle-tier values edged up just 2 percent and bottom-tier values fell 22 percent. Thus, while the housing recovery has reached much of the country, neighbor-hoods hit especially hard by the crash and the recession are still struggling to rebound.

According to CoreLogic data, the broad uptick in home prices reduced the number of homeowners underwater on their mortgages from 5.3 million in the fourth quarter of 2014 to 4.3 million in the fourth quarter of 2015. While this is a far cry from the 12.1 million peak at the end of 2011, the share of homeowners with high loan-to-value (LTV) ratios remains elevated. Of the homeowners that were still underwater last year, 20 percent had LTV ratios of 100–105, 44 percent had LTVs of 105–125, and 38 percent had LTVs of 125 or higher. However, another 1 million homeowners had less than 5 per-cent equity at the end of 2015, leaving them at risk if home prices decline.

While the national picture has brightened considerably, pock-ets of mortgage distress remain. Among the 50 largest metro areas, the share of mortgaged owners with negative equity was under 2 percent in Austin, Houston, Portland, San Jose, and San Antonio, but over 19 percent in Las Vegas, Miami, Orlando, and Tampa.

HOUSING AND THE ECONOMYResidential fixed investment (RFI), which includes both new construction and homeowner improvement spending, is a critical component of the economy. In 2015, RFI generated $600 billion or 3.3 percent of gross domestic product (GDP), a significant increase from the all-time low of 2.4 percent hit in 2011 (Figure 12). RFI also contributed 10 percent or 0.35 per-centage point to real GDP growth last year, providing a lift far exceeding its relative size in the economy.

On the improvements side, rising prices for rental proper-ties have stimulated a surge of spending. Total spending on improvements, maintenance, and repairs to rental units rose nearly 10 percent in 2014, to just under $60 billion. Most improvement expenditures on multifamily properties are for replacement projects, such as building system upgrades or new roofing or flooring. While spending in this category has increased since the downturn, maintenance and repair expen-ditures have been essentially flat, leaving room for additional investment in the rental stock.

Meanwhile, homeowner improvement spending accounted for just over a third of residential construction spending last year, down from about half during the worst of the housing crisis in 2011. Before the crash, homeowners devoted about 40 percent of their remodeling budgets to replacement projects, about 40 percent to discretionary projects such as kitchen and bath remodels, and the remaining 20 percent to property improve-ments and disaster repairs. After cutting back sharply when the crisis hit, homeowners are now undertaking more discretionary projects. At last measure in 2013, discretionary spending was

Source: JCHS tabulations of S&P/Case-Shiller House Price Indexes.

● Los Angeles ● Denver ● Las Vegas ● Detroit ● US Average

300

250

200

150

100

0

2000 2003 2006 2009 2012 2015

Although up Substantially in Most Metros, Home Price Appreciation in Some Markets Still LagsIndexed House Prices

FIGURE 11


Source: JCHS tabulations of BEA, National Income and Product Accounts.

7

6

5

4

3

2

1

0

200419961995 200019991997 1998 2002200119911990 1993 19941992 2006 2008 201120102003 2005 2007 2009 2013 20152012 2014

The Housing Sector Is Gradually Returning to Its Traditional Share of the EconomyResidential Fixed Investment as a Share of GDP (Percent)

FIGURE 12

Average

up 6.9 percent from 2011 and back above 30 percent of total homeowner improvement expenditures. As home prices rise and owners continue to build equity, they are likely to take on more of these big-ticket projects.

Rising property values should also generate housing wealth effects. Historically, as home equity grows, households increase their consumer spending by several cents on the dollar. Estimates from Moody’s Analytics, however, suggest that the impact of housing wealth (as measured by the value of the national housing stock) on retail sales declined by half after the housing bubble burst. But this same study also found that the housing wealth effects in metro areas where home prices were back to pre-crisis peaks in 2014 were about 2.5 times those in metros where home prices had not fully recovered. With home prices now strengthening in most markets, housing wealth effects should thus provide a lift to both consumer spending and the economy.

THE OUTLOOKA number of positive trends—continued strong gains in mul-tifamily construction, growing momentum in single-family construction, increases in new and existing home prices and sales, and further reductions in mortgage distress—made 2015 a year for cautious optimism. In fact, NAHB’s measure of home-builder confidence ended 2015 at its highest level since 2005. Homeowners are also feeling encouraged, with nearly half of all respondents to the latest Fannie Mae National Housing Survey believing it was a good time to sell—a sign that for-sale inven-tories may be set to expand. All of these indicators, along with measures of income and employment growth, will be important to watch in 2016 because of their direct implications for house-hold growth and housing demand.

DEMOGRAPHIC DRIVERS


UPTURN IN HOUSEHOLD GROWTH After years of weakness, growth in the number of US house-holds has begun to strengthen. According to the Housing Vacancy Survey, household growth averaged just 625,000 annually in 2007–2013. The pace of growth has now picked up, rising from 653,000 in 2013 to 1.0 million in 2014 and then to 1.3 million in 2015—marking the largest single-year increase in a decade. Although monthly counts from this survey show household growth moderating in early 2016, persistently tight housing markets amid rising construction volumes suggest that household formations are still on the increase.

Other major surveys also point to an uptick (Figure 13). The American Community Survey put household growth at 968,000 at last measure in 2014, up from 652,000 on average in 2007–2013. The Current Population Survey, a much more volatile measure, indicates that household growth averaged 1.1 million over the past two years, up modestly from the 867,000 average annual increases in 2007–2013 but far higher than the 380,000 average annual increases posted in 2009 and 2010.

MILLENNIALS COMING OFF THE SIDELINESThe recent slowdown in household growth was remarkable given that it corresponded with the coming of age of the millen-nials (born 1985–2004), the largest generation in history. Over the past 10 years, the number of adults under age 30 increased by roughly 5 million but the number of households in that age group rose by just 200,000. Indeed, if young adults headed households at the same rates that they did in 2005, there would be 1.7 million more households in this age group today.

Over the next decade, however, the aging of the millen-nial generation will be a boon to household growth (Figure 14). Household headship rates rise from about 25 percent for adults in their early 20s to about 50 percent for those in their 30s. As they move further into these age groups, millennials are expected to form well over 2 million new households each year on average, raising their numbers from 16 million in 2015 to a projected 40 million in 2025.

Household growth, the primary

driver of housing demand, is

recovering. Incomes, immigration,

and domestic migration are

on the rise, and the millennial

generation is poised to form

millions of new households over

the next decade. While the baby

boomers will be less active in

the homebuying market as they

approach retirement age, they

will give a boost to improvement

spending as they invest in

projects that allow them to

remain in their homes. With the

population aging and minorities

driving most of the growth in

households, the demand for

housing will become increasingly

diverse.


The recent upturn in household growth does not reflect any rebound in headship rates among young adults. Indeed, rates of living with roommates remain slightly elevated, and the share of young adults living with their parents continues to rise. The American Community Survey indicates that the share of 25–34 year-olds living in their parents’ homes rose from 17 percent in 2008 to about 22 percent in 2014, and more recent Current Population Survey data show further increases in 2015.

Adults living with parents are mainly in their 20s, with the shares declining sharply from 50 percent among adults aged 20–24, to 27 percent among those aged 25–29, to 15 percent among those aged 30–34. According to the Fannie Mae National Housing Survey, the major reasons for living with parents dif-fer somewhat across these age groups, but commonly involve minimizing housing expenses. For example, adults in their early 20s are more likely to be unemployed and live with their parents because they are still enrolled in school. In contrast, adults in their mid-20s to early 30s are more likely to be out of school and employed, but still living with parents for the cheap housing and to build their savings while working.

High housing costs are clearly a barrier to living independently for many younger adults. In the 100 largest metros, household headship rates for 25–29 year-olds are significantly lower in areas where housing is least affordable (as measured by renter cost-burden rates). Indeed, headship rates among this age group in the 25 least affordable metros are a full 10 percentage points lower than those in the 25 most affordable metros. Least-affordable metros include high-cost areas such as New York and Los Angeles, but also Philadelphia, Fresno, and Lakeland, where rents are more moderate but still high relative to incomes.

GROWING INCOMES BUT GROWING INEQUALITYLow incomes also prevent young adults from living on their own, so the recent pickup in income growth is good news for housing demand. With employment rising for the 67th consecu-tive month in April 2016, job growth has slowly translated into measurable income gains. Real median income for all workers age 15 and over edged up 1.0 percent in 2014, the third year of increases. Young adults made even more progress, with a 2.3 percent increase for workers aged 25–34 and 4.1 percent for workers aged 35–44 (Figure 15). While back above recent lows, the real median personal incomes for these age groups are still 9–18 percent below previous peaks.

Household headship rates among 25–34 year-olds rise sharply with income, starting from 40 percent for those earning less than $25,000, to 50 percent for those earning $25,000–49,999, to 58 percent for those earning $50,000 or more. This strong link suggests that much of the recent decline in household forma-tions among young adults is income-related. A JCHS analysis of Current Population Survey data confirms this fact, finding that if the income distribution among 25–34 year-olds had remained

Note: American Community Survey data are only available through 2014.Source: JCHS tabulations of US Census Bureau survey data.

● American Community Survey ● Housing Vacancy Survey ● Current Population Survey

2000–2007 2007–2013 2013–2015

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0

Major Census Surveys Confirm the Pickup in Household GrowthAverage Annual Household Growth (Millions)

FIGURE 13

Source: JCHS tabulations of US Census Bureau, United States Population Estimates and 2014 Population Projections.

● 2005–2015 ● 2015–2025

20–24 25–29 30–34 35–39 40–44

4

3

2

1

0

-1

-2

-3

Age Group

Over the Next Ten Years, the Aging of the Millennial Generation Will Boost the Population in Their 30sPopulation Growth (Millions)

FIGURE 14


constant in 2005–2015, their headship rates would have dropped only half as much. Further income gains among young adults should therefore help to reverse some of the decline in their headship rates.

Meanwhile, the real median income of US households inched up 1.2 percent in 2014, the second consecutive year of growth. At $53,700, however, real median income was still 6 percent below the 2007 peak and lower than any pre-recession level dat-ing back to 1996. Moreover, income disparities have increased sharply over the past few decades, with the average real income of households in the bottom decile down 18 percent in 1980–2014 (from $7,700 to $6,300) and that of households in the top decile up 66 percent (from $154,000 to $256,000). Average real incomes for the middle two deciles grew a modest 6 percent over this period. As a result, the average top-decile household now makes 40 times the income of the average household in the bottom decile and 4.7 times that of the average household in the middle deciles.

Reflecting the uneven growth in incomes, households earning under $25,000 per year were the fastest-growing segment in 2005–2015. Indeed, their numbers rose by 21 percent over the decade. As a result, this low-income group accounted for 44 percent of the nation’s net growth in households.

DISPARITIES IN HOUSEHOLD WEALTHAlong with income, wealth is increasingly concentrated in the hands of the few and the numbers of households with little or no assets continue to increase. The Survey of Consumer Finances indicates that the share of wealth held by households in the top income decile jumped from 53 percent in 1992 to 62 percent in 2013. Meanwhile, the share of wealth held by households in the bottom half of the income distribution declined from 15 percent to 10 percent. As a result, the number of households with less than $25,000 in real net wealth rose from about 30 million to more than 43 million over this period, including nearly 20 mil-lion with wealth of $1,000 or less (Figure 16).

Over three-quarters of the households with less than $25,000 in wealth are renters, underscoring the close link between wealth and homeownership. At last measure in 2013, the typical home-owner had $195,500 in net household wealth while the typical renter had just $5,400. Households with traditionally low home-ownership rates—minority and low-income households—are therefore at a significant disadvantage. Indeed, the substantial white-minority homeownership gap has left the median net household wealth of blacks ($11,000) and Hispanics ($13,700) at roughly one-tenth that of whites ($134,200). And for those able to make the transition to homeownership, home equity makes up a disproportionately large share of net wealth. In 2013, home equity accounted for more than 80 percent of the net wealth of low-income homeowners and well over 50 percent of the net wealth of minority homeowners.

Age Group ● 25–34 ● 35–44 ● All Adults

Note: Data are for adults age 15 and over. Source: JCHS tabulations of US Census Bureau, Current Population Surveys.

40

38

36

34

32

30

28

26

24

22

201996 1998 2000 2002 2004 2006 2008 2010 2012 20141994

After Years of Decline, Real Incomes for Young Adults Are Finally on the RiseMedian Income Per Capita (Thousands of 2014 dollars)

FIGURE 15

Note: Dollar values are adjusted to 2013 dollars using the CPI-U for All Items.Source: JCHS tabulations of US Federal Reserve Board of Governors, Surveys of Consumer Finances.

45

40

35

30

25

20

15

10

5

0

1992 1995 1998 2001 2004 2007 2010 2013

Millions of Households Have Little or No WealthHouseholds (Millions)

FIGURE 16

Household Net Worth● $5,001–25,000 ● $1,001–5,000 ● $1–1,000 ● Zero or Negative


But for many young adults, low wealth remains an obstacle to homebuying. In 2013, renters aged 25–34 had median net wealth of $4,850 and cash savings of $1,030, well below the downpayment needed for today’s median-priced home. Renters aged 35–44 were not much better off, with median net wealth of $7,900 and cash savings of $510. Given the large discrepancy in wealth between owners and renters, the inability to access homeownership may further divide the haves and the have-nots.

REBOUND IN IMMIGRATIONImmigration, another major driver of household growth and housing demand, is starting to pick up steam. From a low of 704,000 in 2011, net international immigration climbed to an estimated 1.15 million last year. This latest influx has changed the mix of new residents, with today’s immigrants more likely to be Asian than Hispanic (Figure 17). This shift largely reflects a falloff in immigration from Mexico as well as an increase in outmigration from the US to Mexico. The Pew Research Center reports that the number of Mexican immigrants fell from 2.9 million in 1995–2000 to 870,000 in 2009–2014, while emigration of US residents to Mexico increased from 670,000 to 1.0 million, resulting in a net population loss of 140,000.

The changing mix of immigrants has direct implications for housing demand. Asian immigrants generally have higher incomes, higher homeownership rates, and higher levels of educational attainment than Hispanic immigrants. In 2014, foreign-born Asian households aged 25–44 had a median income of $82,000, or more than twice the $38,000 median

income of similarly aged, foreign-born Hispanic households. In addition, the homeownership rate for foreign-born Asians was 57 percent, 15 percentage points higher than for foreign-born Hispanics. Asian immigrants also had slightly fewer children and were more likely to settle in the Northeast and Midwest than Hispanic immigrants.

Immigrants have been an important source of household growth for decades. According to the Current Population Survey, the foreign-born contributed just over a third of the increase in households in 1994–2015, or about 450,000 house-holds per year. Indeed, just when the large baby-boom gen-eration was moving out of the prime household formation years, immigrants bolstered the ranks of the smaller gener-ation-X population (born 1965–1984), changing the composi-tion of that generation and stabilizing housing demand. The foreign-born thus accounted for nearly a fifth of household heads aged 30–49 in 2015.

Given the strong inflows of Hispanic immigrants in the late 1990s and early 2000s, the minority share of the gen-X popu-lation now stands at 41 percent. By comparison, the minority share of millennials is slightly higher at 45 percent, while that of the baby boomers is just 29 percent. Going forward, as the millennials replace the gen-Xers among households in their 30s and 40s, immigrants will fuel additional need for housing, adding to the strong demand expected from what is already the largest, most diverse generation in history.

RESIDENTIAL MOBILITY TRENDSResidential mobility rates, or the share of the population that changes homes in a given year, have trended downward since the mid-1990s. Mobility rates are highest for adults under age 25 (39 percent) and decline steadily across age groups, falling to just 3 percent for households age 75 and over. The aging of the baby-boom generation and increases in longevity have thus reduced the overall mobility rate by raising the share of the population that is least mobile.

But mobility rates for all age groups have also slipped in recent years. In fact, the largest declines have been among households under age 25, with rates now 15 percentage points below those in 2000. By comparison, rates are down 5 percentage points for 25–34 year-olds, 3 percentage points for 35–44 year-olds, and 2 percentage points for 45–54 year-olds. Several factors have contributed to this trend, including lower household forma-tion rates among young adults and lower homebuying activity among older adults.

Less frequent moves among renters are also a key factor. When the housing downturn began in 2005, the sharpest drop in mobility rates was among homeowners kept in their current homes by the collapse of house prices and limited access to credit. Homeowner mobility rates fell from 6.3 percent to 4.1

Note: Whites, blacks, and Asians are non-Hispanic. Hispanics may be of any race.Source: JCHS tabulations of US Census Bureau, 2014 American Community Survey 1-Year Estimates.

● 1990–2009 ● 2010–2014

Hispanic AsianBlack White

700

600

500

400

300

200

100

0

Asians Are Leading the Current Wave of ImmigrationAverage Annual Immigration (Thousands)

FIGURE 17


percent over the ensuing five years before stabilizing. At the same time, renter mobility rates slipped by 1.4 percentage points in 2005–2010, but then dropped 3.8 percentage points in 2010–2015. Contributing to this slowdown were historically low household formation rates (implying fewer moves per capita into rentals) and longer stays in current units (reflecting in part fewer transitions into homeownership).

Still, domestic migration (state-to-state moves within the coun-try) increased in 2015, restoring the long-term flow of popula-tion into the South and West (Figure 18). After falling 48 percent in 2007–2013, net population growth in the South rebounded to a post-recession high of 444,240 last year, led by the Sunbelt states of Florida, North Carolina, and Texas. Meanwhile, net population losses in the Northeast and Midwest—led by Illinois and New York—increased to their pre-recession levels.

One of the most noteworthy changes in mobility patterns after the Great Recession was slower population growth in suburban areas and faster population growth in urban areas. Weaker migration to the Sunbelt was one factor, given that metros in that region are much less compact than in the North. The sharp falloff in single-family construction also contributed since much of that type of housing is developed in suburban and exurban locations. As domestic migration resumes and single-family construction picks up, however, suburban growth rates are likely to rebound. In fact, a 2015 analysis by the Brookings Institution indicates that the turnaround has already begun, with population growth in suburban counties exceeding that in urban core counties for the first time since 2009.

But there is no question that the recent strength of urban population growth is driven in part by growing demand for city living. However, the 2015 Demand Institute Consumer Housing Survey indicates that the majority of US households prefer to eventually settle in suburban or exurban communities. Among households expecting to move in the next five years, 69 percent intended to live outside of city centers. This share rises to 78 percent of those planning to move into homes they own. Even among respondents under age 35, fully 63 percent of future movers intended to live outside of a city center, including 71 percent of those planning to own.

THE OUTLOOKGrowth in the adult population will support significant house-hold growth over the next decade and beyond. According to pre-liminary JCHS projections, demographic forces alone will drive the addition of more than 13 million households in 2015–2025. Much of this growth will occur among the retirement-aged population, with the number of households age 70 and over projected to soar by over 8 million, or more than 40 percent. These increases will lift the share of older households from 16 percent in 2015 to about 21 percent in 2025.

The aging of the population will have profound impacts on hous-ing demand. First, the growing share of older households means further declines in residential mobility and housing turnover, potentially putting the already tight market for existing homes under additional pressure. Second, as they age in place in great-er numbers, older households will not only contribute a larger

Source: JCHS tabulations of US Census Bureau, United States Population Estimates.

● Northeast ● Midwest ● South ● West

600

500

400

300

200

100

0

-100

-200

-300

-4002005 2007 2009 2011 2013 2015

Domestic Migration Is Returning to Historical Patterns of Growth and LossNet Domestic Migration (Thousands)

FIGURE 18


share of remodeling spending, but will also increase demand for different types of projects, such as accessibility improvements. Third, the older households that do move will likely seek units that are smaller and less costly to maintain—the same types of housing young adults want to rent or buy as their first homes. And finally, the number of older single persons living alone will climb, implying a significant increase in the need for in-home healthcare and supportive services.

Meanwhile, the millennials will have a growing presence in housing markets as the younger members of this large genera-tion enter adulthood and older members move into the prime first-time homebuying years. While their aspirations for hous-ing do not differ significantly from those of previous genera-tions, millennials have come of age in an era of lower incomes, higher rents, and more cautious attitudes towards credit and homeownership, conditions that are likely to affect their con-sumption of housing for years to come.

The racial and ethnic diversity of this huge generation will drive up the number and share of minority households. Indeed, minorities are expected to account for roughly 75 percent of household growth over the next 10 years. The growing minor-ity share in housing markets is likely to boost demand for units that accommodate multigenerational households, given that young minority adults are more likely than young white adults to live with their parents and older minority adults are much more likely than older white adults to live with their children. More importantly, however, rapid growth in minority house-holds brings new urgency to the need to reduce white-minority gaps in household income, wealth, and homeownership. Failure to make progress in this realm could have increasingly large impacts on the shape of future housing demand.

HOMEOWNERSHIP


HOMEOWNERSHIP RATE DECLINESThe decade-long slide in homeownership is unprecedented in American history, with the national rate down more than 5 percentage points from the 69.0 percent peak in 2004, to just 63.7 percent in 2015 (Figure 19). The persistent decline reflects the lack of growth in the number of homeowner households at a time of robust growth in the number of renter households. According to the Housing Vacancy Survey, the number of renter households hit 42.6 million last year, an increase of 1.4 million from 2014 and some 9.3 million from 2004. Meanwhile, the number of homeowner households slipped to 74.7 million in 2015, down 87,000 from 2014 and up just 431,000 from 2004.

While homeownership rates for households of all ages and races/ethnicities have fallen, the size of the declines varies across groups. The largest drop has been among 35–44 year-olds, with rates dropping nearly 11 percentage points from 69.2 percent in 2004 to 58.5 percent in 2015. By comparison, the homeownership rate fell about 8 percentage points among households under age 35, about 7 percentage points among households aged 45–54, about 6 percentage points among households aged 55–64, and just 2 percentage points among households aged 65 and over. As a result, homeownership rates for all but the oldest age group are now lower than in 1994. In fact, the national rate remains near its 1994 level only because of the overall aging of the population, which means that increasing numbers of households are now in the age groups when homeownership rates are highest.

Following these declines, the gap between black-white home-ownership rates widened, while the gap between Hispanic-white rates narrowed slightly. The share of white households that owned homes in 2015 was down 4.0 percentage points from the 2004 peak, to 71.9 percent. Meanwhile, the homeowner share of all minority households fell 4.3 percentage points, end-ing 2015 at 46.7 percent. Over this same period, the share of black households owning homes dropped 6.7 percentage points, to 43.0 percent, while the share of Hispanic households owning homes declined 2.5 percentage points, to 45.6 percent.

With mortgage credit still tight

and foreclosures relatively high,

the national homeownership rate

continued to trend downward in

2015. Although low inventories of

homes for sale are keeping prices

on the rise, homeownership in

many metropolitan areas remains

affordable by historical standards

and most Americans continue to

believe that owning a home is a

sound financial investment. The

ongoing recovery in the economy

may reinvigorate demand for

homeownership, although how

quickly a rebound might occur

remains an open question.


FORECLOSURES BACKLOG AND SLUGGISH HOMEBUYINGThe overhang of foreclosures has contributed to the continued slump in homeownership. Although on the decline, distressed sales are still well above pre-crisis levels (Figure 20). According to CoreLogic data, the total number of foreclosure completions, short sales, and deed-in-lieu of foreclosure transactions for one- to four-family properties averaged 55,900 per month in 2015. This figure is down from a peak of 120,200 per month in 2010,

but only a small improvement from the 67,100 monthly aver-age in 2014. By comparison, foreclosure-related sales ran at just 19,900 per month from 2000 to 2005.

However, foreclosures are likely to continue to recede in the coming years. The Mortgage Bankers Association reports that the inventory of properties in the foreclosure process totaled 688,000 units in the fourth quarter of 2015, a significant improvement over the 929,000 units in 2014 and the high of 2.1 million in 2010. This is the smallest inventory since 2007, with declines occurring in all but three states (Delaware, Massachusetts, and Rhode Island) last year. In addition, new foreclosure starts and loan delinquencies also fell in 2015. Only 140,000 foreclosures were started in the fourth quarter of 2015, a drop of 48,000 from a year earlier. The share of owners with mortgages that were seriously delinquent on their loans (90 or more days past due or in foreclosure) stood at 3.4 per-cent at the end of 2015, down from 4.5 percent at the end of 2014 and a high of 9.7 percent in 2009.

With foreclosures on the decline, the future trajectory of the homeownership rate depends largely on the speed of recov-ery in home purchases, particularly among first-time buyers. According to NAR data, the share of sales that were first-time purchases dipped from 33 percent in 2014 to 32 percent in 2015, down from 40 percent in 2003–2005. In addition, Current Population Survey data indicate that in 2015, only 2.7 percent of US households moved into homes purchased within the last year, compared with 4.7 percent in 2000 (Figure 21). While this measure has trended upward in each age group since 2013, the speed of recovery in coming years remains uncertain.Note: Data are four-quarter rolling averages.

Source: JCHS tabulations of US Census Bureau, Housing Vacancy Surveys.

● Homeownership Rate (Left scale) ● Homeowners (Right scale)

70

69

68

67

66

65

64

63

62

61

60

100

95

90

85

80

75

70

65

60

55

50

1985 1990 1995 2000 2005 2010 2015

With No Growth in the Number of Owners, the National Homeownership Rate Fell Again in 2015Percent

FIGURE 19

Millions

Source: JCHS tabulations of CoreLogic data.

160

140

120

100

80

60

40

20

0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Distressed Sales Have Declined But Remain above Pre-Crisis LevelsMonthly Foreclosure Completions, Deed-in-Lieu Transactions, and Short Sales (Thousands)

FIGURE 20


The slowdown in homebuying does not appear to be due to household attitudes or perceptions of risk, as most Americans still believe that homeownership is a sound financial choice. According to a 2015 Demand Institute survey, 78 percent of household heads agreed with the statement “I think home-ownership is an excellent investment,” while only 6 percent disagreed. Although renters were somewhat less favorable, more than 67 percent agreed that homeownership is an excel-lent investment, and just 10 percent disagreed. Younger renter households were especially positive on this point, with 75 per-cent of renters below age 40 agreeing about the financial value of homeownership.

A recent Joint Center analysis of the University of Michigan’s Panel Study of Income Dynamics data illustrates the wealth-building potential of sustained homeownership. For example, the median increase in real net wealth among households that remained homeowners between 1999 and 2013 was $91,900, including a $37,100 gain in home equity. Households that bought homes between 1999 and 2009 and were able to sustain homeownership through 2013 also saw significant growth in net wealth of $85,400, including a $46,000 increase in home equity. To be sure, homeownership is not without risks. The median household that bought a home in 1999–2009 but did not continue to own a home through 2013 lost $2,800 in net wealth. Meanwhile, the median household that rented throughout this period saw no change in net wealth.

AFFORDABILITY OF HOME PRICES While home prices have rebounded from their 2011 lows, they are still affordable by historical standards. The real median sales

price of existing homes climbed 6.6 percent in 2015 to $222,400, while the real median price of new single-family homes rose 4.7 percent to $296,400. Despite this increase, the NAR Housing Affordability Index—comparing median household income to the mortgage payment on a median-priced home—suggests that affordability declined only slightly last year.

Low mortgage interest rates helped, with the average rate on 30-year fixed-rate loans holding below 4.0 percent for most of 2015 and standing at 3.6 percent in April 2016. These low rates kept the increase in principal and interest payments for a median-priced home in 2014–2015 to just 2.6 percent, lifting the median payment to $834 (Figure 22). Using a broader measure of households’ total monthly expenditures on housing and utilities from the American Community Survey, the real median monthly housing cost for homeowners who moved into their units within the previous year rose 4.8 percent in 2013–2014, to $1,203.

At the metropolitan level, however, affordability varies dra-matically. At one extreme, the ratio of median home price to median household income in the Rockford (Illinois) metropoli-tan area was just 1.8 in 2014, compared with 3.9 for the nation as a whole. But at the other extreme, the price-to-income ratio in Honolulu was 9.1 in 2014. This range illustrates the sharply different market conditions that would-be homebuyers face depending on their location.

STUDENT LOAN DEBT AND DOWNPAYMENT PRESSURESAlthough home prices remain generally affordable, rising stu-dent loan debt has eroded the amount of income that house-holds have to spend on a home purchase. According to the

Notes: Home purchases are equal to the number of homeowners that moved in the preceding year. Data are three-year trailing averages. Source: JCHS tabulations of US Census Bureau, Current Population Surveys.

Age Group ● Under 35 ● 35–49 ● 50–64 ● 65 and Over ● All

8

7

6

5

4

3

2

1

0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Homebuying Activity Is Still Depressed But No Longer DecliningShare of Households that Purchased Homes in the Previous Year (Percent)

FIGURE 21


Survey of Consumer Finances, the share of all US households with outstanding student loan debt increased from 12 percent in 2001 to 20 percent in 2013. At the same time, the median outstanding loan balance rose from $10,500 to $17,000, with 36 percent of borrowers in 2013 owing more than $25,000 and 17 percent owing more than $50,000.

While households of all ages have taken on additional student loan debt, the largest increase has been among the young. Some 39 percent of households aged 20–39 carried student loan debt in 2013, compared with 19 percent of households aged 40–59 and 5 percent of households age 60 and over (Figure 23). Among this older group, outstanding debt may be a combination of loans taken out to pay for their children’s education and their own mid-life educational costs.

For young renters that want to buy homes, student loan debt can add considerably to the debt-to-income ratio that lenders use to determine eligibility for mortgage loans. Among 20–39 year-olds with student loan debt payments, the mean loan pay-ment in 2013 ranged from 4 percent of income among renters in the highest income quartile to 15 percent of income for rent-ers in the lowest income quartile. These payments are on top of student loan borrowers’ other non-housing debt payments, which consumed on average another 4 percent of income for renters in the highest income quartile and 7 percent of income for renters in the lowest income quartile.

Accumulating a downpayment presents an additional chal-lenge. The 2013 Survey of Consumer Finances indicates that

12 percent of renter households had no savings in transac-tion or retirement accounts or other financial instruments. Among the other 88 percent of renter households, the medi-an value of all financial assets was just $3,000. By compari-son, a 5 percent downpayment on a median-priced existing home in 2015 was $11,100.

The Federal Housing Administration (FHA), which offers loans with downpayment requirements as low as 3.5 percent, has traditionally been a critical source of mortgage credit for house-holds unable to put large amounts down on a home purchase. Fannie Mae and Freddie Mac also lowered their downpayment requirements from 5 percent to 3 percent in 2015, introducing loan products that target first-time homebuyers who otherwise meet underwriting criteria and complete pre-purchase home-ownership education and counseling. Fannie Mae and Freddie Mac also strengthened their partnerships with state housing finance agencies, which are another vital source of downpay-ment assistance and related first-time homebuyer programs. Both efforts may help to reduce the obstacles that first-time homebuyers face in qualifying for mortgages, particularly in high-cost markets where downpayment thresholds are a sig-nificant barrier.

TIGHT MORTGAGE MARKET CONDITIONS The number of first-lien mortgage originations for owner-occupied home purchases increased only incrementally from its 2011 low, reaching 2.8 million in 2014. While originations are still below their 2000 levels, Fannie Mae and Freddie Mac both

Notes: Incomes are adjusted for inflation using the CPI-U for All Items. Home prices are adjusted using the CPI-U for All Items less shelter. Monthly mortgage payments include principal and interest, and assume a 30-year, fixed-rate mortgage with a 20% downpayment.Source: JCHS tabulations of NAR, Single-Family Existing Home Prices; Moody’s Economy.com, Median Family Incomes; and Freddie Mac, Primary Mortgage Market Surveys.

● Monthly Mortgage Payment on Median-Priced Existing Home (Left scale)

● Median Home Price (Right scale) ● Median Family Income (Right scale)

1,600

1,400

1,200

1,000

800

600

400

200

0

320,000

280,000

240,000

200,000

160,000

120,000

80,000

40,000

0

1985 1990 1995 2000 2005 2010 2015

Despite Rising Prices, Mortgage Payments Have Remained Relatively Low 2015 Dollars

FIGURE 22


project modest growth in home purchase mortgage volumes to continue in 2016 and 2017.

Meanwhile, mortgage credit remains tight for borrowers unable to meet strict underwriting standards. The Urban Institute’s Housing Credit Availability Index indicates that credit conditions changed very little in 2015 and were only slightly looser than at their post-recession trough. Similarly, the MBA’s Mortgage Credit Availability Index does not show a clear trend in 2015 and confirms that market conditions remain relatively tight.

As a result, lending to households with less than perfect credit histories has fallen off. CoreLogic data indicate that loans to homebuyers with observed credit scores below 700 declined from 33 percent of first-lien mortgages in 2010 to just 27 percent in 2014. This tightening of standards has, however, kept cumu-lative default rates on Fannie Mae and Freddie Mac’s 2011–2014 loans well below those for any prior loan vintage from 1999 to 2010. Taken together, these trends suggest that recent growth in the number of conventional mortgage originations has been primarily to low-risk borrowers.

Tight credit conditions limit access to homeownership, particu-larly for low-income and minority households. Home Mortgage Disclosure Act (HMDA) data show that the share of mortgage

loan originations to low- and moderate-income homebuyers fell from 36 percent in 2010 to 27 percent in 2014. Over this same period, the share of originations to black homebuyers edged down from a modest 6 percent to 5 percent, while the share to Hispanic homebuyers remained steady at about 8 percent.

In 2015, FHA, Fannie Mae, and Freddie Mac all took steps to expand mortgage credit availability. In addition to introducing lower downpayment options, both Fannie Mae and Freddie Mac updated and clarified their origination and servicing standards, as well as policies for repurchase requests, in an effort to reduce the credit overlays applied by many lenders. FHA took similar steps and also lowered its mortgage insurance premium from 1.35 percent to 0.85 percent. Despite these and other moves, however, measures of mortgage credit availability showed that conditions remained tight in the first quarter of 2016.

CHANGES IN MORTGAGE ORIGINATION CHANNELS The weakness in mortgage originations in 2015 was accompa-nied by changes in the regulatory environment resulting from the rollout of Dodd-Frank Act provisions and other rulemaking. These changes, along with recent enforcement actions, may have dampened lending activity as lenders adjust to the new standards.

The Ability to Repay rule (also known as the Qualified Mortgage rule), which took effect in January 2014, requires mortgage loan originators to collect more income documentation and verify applicants’ ability to afford new loans. Although noting slight reductions in the share of high-priced loans following imple-mentation, a Federal Reserve Board analysis of HMDA data concluded that the new rule “did not materially affect the mort-gage market in 2014.” In the future, however, the rule may have larger impacts if credit conditions loosen sufficiently to increase lending to higher-risk borrowers.

Building on these changes, the Consumer Financial Protection Bureau’s “Know Before You Owe” rule was rolled out in October 2015, revising the disclosure documents that lenders must pro-vide borrowers. Lenders have argued that the new disclosure forms raise origination costs and lengthen the time required for some closings. However, it is still too early to know whether any increase in origination costs will dissipate once lenders adjust to the new standards, or to determine whether the new disclosure forms substantially improve borrowers’ experiences.

At the same time that the regulatory environment is changing, the types of institutions originating and funding loans are also undergoing a shift. Between 2010 and 2014, independent mort-gage companies continued to grow their market share from 35 percent of home purchase mortgage originations to 47 percent (Figure 24). In contrast, the bank share declined steadily from 50 percent to 40 percent over this same period. Meanwhile, Fannie Mae and Freddie Mac remain the primary funding source for

Note: Households not yet in repayment have student loans in deferral due to schooling, military service, emergency hardship,

or other reasons.

Source: JCHS tabulations of Federal Reserve Board of Governors, Surveys of Consumer Finances.

40

35

30

25

20

15

10

5

0

20–39

2001 2013 2001 2013 2001 2013

40–59 60 and Over

Age Group

● Not Yet in Repayment ● 3 and Under ● 4–7 ● 8–13 ● 14 and Over

Student Loan Payments as Percent of Income

Many Younger Households Have to Devote a Significant Share of Income to Student Loan PaymentsShare of US Households (Percent)

FIGURE 23


mortgage originations, with private-label securities account-ing for less than 5 percent of gross issuance of new mortgage backed securities.

THE OUTLOOKIn the short term, tight credit conditions, the limited supply of homes for sale, and relatively high foreclosure volumes may continue to push down the national homeownership rate. Over the long term, however, demographic patterns, household atti-tudes, and economic conditions are likely to play larger roles in shaping the market.

The aging of the large millennial generation has the potential to produce millions of new homeowners in the coming years. In fact, most Americans believe that homeownership is not only desirable but also attainable (Figure 25). A 2015 Demand Institute survey found that 83 percent of respondents expected to own homes in the future. Among renters, 52 percent expected to own homes, including 28 percent who anticipated buying a home with their next move, and 24 percent who expected to buy “someday.” Moreover, especially large shares of younger renters expect to become homeowners, including 85 percent of those under age 30 and 69 percent of those aged 30–39.

The ability of these households to buy homes will depend on future economic, housing, and credit conditions. While these factors are difficult to predict, slowing foreclosures and the relative affordability of homeownership suggest that the owner-occupied housing market is likely to recover. The coming years will be instructive about the extent to which improving econom-ic conditions translate into broader demand for homeowner-ship, as well as into increases in the supply of homes accessible to entry-level homebuyers.

● 2000 ● 2005 ● 2010 ● 2014

80

70

60

50

40

30

20

10

0Banks, Credit Unions, Affiliates Independent Mortgage Companies

Independent Mortgage Companies Have Increased Their Share of the Home Purchase Loan MarketShare of Originations (Percent)

FIGURE 24

Notes: Originations include all first-lien home purchase mortgages for one- to four-family, owner-occupied, site-built homes. Affiliates include mortgage companies owned by a bank, credit union, or its parent company.Source: N. Bhutta, J. Popper, and D. Ringo, The 2014 Home Mortgage Disclosure Act Data, Federal Reserve Bulletin 101(4), November 2015.

Source: JCHS tabulations of The Demand Institute, 2015 Consumer Housing Survey data.

100

80

60

40

20

0Under 30 30–39 40–49 50–59 60–69 70 and Over

Age Group

● Do Not Expect to Buy a Home ● Expect to Buy a Home in Next Move

● Expect to Buy a Home Someday ● Currently Own Home

The Vast Majority of Households Either Own Homesor Expect to in the FutureShare of Survey Respondents (Percent)

FIGURE 25

RENTAL HOUSING


A VITAL RESOURCE FOR A DIVERSE NATIONRental housing serves all types of households in a broad range of communities. In total, about 36 percent of US households—representing nearly 110 million people, including 30 million children—lived in rentals last year. While more than half of central city households rented their housing, the renter shares in suburban communities (28 percent) and in non-metro areas (27 percent) are also large.

Renters are more diverse than homeowners in terms of age, income, and household type (Figure 26). Although young adults are the age group most likely to rent, 34 percent of renter households are headed by an individual age 50 and over and 40 percent by an individual aged 30–49. While more than a third of renter households earn less than $25,000, a sizable and growing number of high-income households also choose to rent for the flexibility and convenience it provides. Families with children, one of the household types most likely to own homes, are increasingly likely to rent. Indeed, families with children make up 31 percent of renters, but only 27 percent of homeowners.

Renters are also more racially and ethnically diverse than homeowners. Minorities and foreign-born households account for half of renter households, compared with just one in four homeowners. The differences are particularly striking among black and Hispanic households, with each group making up 20 percent of renters but less than 10 percent of owners.

A DECADE OF BROAD-BASED DEMAND As measured by the Housing Vacancy Survey, the number of renter households soared by nearly 9 million from 2005 to 2015—the largest increase over any 10-year period on record. Moreover, 2015 marked the largest single-year jump in net new renter households, up 1.4 million, with most of the gains posted in the first half of the year. Renters have thus accounted for all of the net growth in households since 2005 (Figure 27).

Much of the jump in rental demand has come from middle-aged households. Current Population Survey data indicate that the number of renter households in their 50s and 60s rose by 4.3

Rental housing markets across

the country tightened again

in 2015. While multifamily

construction ramped up for the

fifth consecutive year, demand

continued to outstrip supply,

pushing down vacancy rates and

pushing up rents. Although renter

household growth is likely to

slow from its current pace, rental

demand should remain strong

over the coming decade, keeping

markets under pressure—

particularly at the low end.


million in 2005–2015, driven by both the aging of baby-boomer renters and declines in homeownership rates among this age group. Renter households age 70 and over also increased by more than 600,000 over the decade. Meanwhile, households in their 30s and 40s accounted for 3 million net new renters despite the dip in population in this age group. Households under age 30, however, made up only 1 million net new renters, reflecting the steep falloff in headship rates among the millen-nial generation following the Great Recession.

With the overall aging of the US population and the growth in the number of baby-boomer renters, single persons living alone (up 2.9 million) and married couples without dependent children (up 1.6 million) propelled much of the growth in renter households over the past decade. At the same time, though, the number of renters with children—including both couples and single-parent families—rose by 2.2 million.

While demand picked up across households of all incomes, nearly half of the net growth in renters (4.0 million) was among households earning less than $25,000. Even so, the number of new renters earning $50,000 or more increased nearly as much (3.3 million), including 1.6 million households earning $100,000 or more. Top-income households have been the fastest growing segment over the past three years, but still make up only an 11 percent share of all renters.

Notes: Couples include both married and unmarried partners. Families with children include single parents and couples with children under age 18. Data are three-year rolling averages. Source: JCHS tabulations of US Census Bureau, 2013–2015 Current Population Surveys.

100

90

80

70

60

50

40

30

20

10

0

Renters OwnersRenters Owners Renters Owners

Age Group Household Income Household Type

● 70 and Over

● 50–69

● 30–49

● Under 30

● $100,000 and Over

● $50,000–99,999

● $25,000–49,999

● Under $25,000

● All Other

● Single Person

● Couples without Children

● Families with Children

Renter Households Reflect the Full Diversityof US HouseholdsShare of Households (Percent)

FIGURE 26

Source: JCHS tabulations of US Census Bureau, Housing Vacancy Surveys.

2001–2003 2004–2006 2007–2009 2010–2012 2013–2015 2001–2003 2004–2006 2007–2009 2010–2012 2013–2015

Renters Owners

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

-0.4

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

-0.4

Renting Has Surged Over the Past Several Years as Homeownership Has Stalled Average Annual Growth in Households (Millions)

FIGURE 27

JOINT CENTER FOR HOUSING STUDIES OF HARVARD UNIVERSITY 27

Minority and foreign-born households contributed two-thirds of the increase in renters in 2005–2015. Native-born minorities led growth, with 3.9 million households in this group joining the ranks of renters. Foreign-born minorities added another 1.9 million renter households. While minorities and immigrants traditionally drive growth in renter households, the number of native-born white renters also increased by 3.0 million over the past decade.

EVOLUTION OF THE SUPPLYThe single-family housing stock absorbed nearly two-thirds of the decade-long growth in renter households, lifting the single-family share of occupied rentals (including mobile homes) from 34 percent in 2005 to 40 percent in 2015. The sharp increase in single-family rentals resulted from conversions of millions of owner-occupied homes following the housing crash, stemming largely from the wave of foreclosures but also from owners’ reluctance to sell in a depressed market.

In contrast, new construction was responsible for much of the growth in the multifamily stock. Indeed, the number of mul-tifamily starts intended for rent climbed from a low of about 92,000 units in 2009 to 370,000 units in 2015, the highest level since the 1980s. Given the relatively long multifamily develop-ment timeline, starts remain well ahead of rental completions, which increased to 304,000 units last year.

According to the Survey of Market Absorption, new multifamily units have fewer bedrooms on average than those built over the past two decades. More than half of the unfurnished, market-rate rentals in structures with five or more units that were completed in 2014 were either studios or one-bedroom apart-ments—the largest share in history, and well above the 36 per-cent average share in the 1990s and early 2000s. Only 7 percent of apartments added in 2014 had three or more bedrooms, down from about 13 percent in earlier periods. Construction was also more concentrated in urban areas, with 57 percent of comple-tions in the past two years located in principal cities compared with an annual average of 45 percent dating back to 1970.

While newer rentals have always commanded higher prices than older units, the premium for new apartments has risen sharply even as their size has decreased. The median asking rent for a new market-rate multifamily unit built in 2015 was $1,381 per month, more than 70 percent higher than the over-all median contract rent for multifamily apartments. The rent premiums for new studio and one-bedroom apartments were at highs of 90 percent and 78 percent, respectively. The steep rents for new units reflect rising land and development costs, which push multifamily construction to the high end of the market. They are also a measure of the growing demand from high-income renters for luxury apartments.

At the other end of the market, growth in the low-rent supply is largely driven by downward filtering of older units. For example, fully 15 percent of units renting for less than $800 per month in 2013 had rents above this cutoff in 2003 (in inflation-adjusted terms). At the same time, however, many low-rent units were upgraded to higher rents. On balance, filtering increased the sup-ply of units renting for under $800 by just 4.6 percent between 2003 and 2013, a gain that was more than offset by the per-manent loss of 7.5 percent of similarly priced units (Figure 28). Factoring in other changes to the stock, the number of low-cost units rose only 11.2 percent over the decade—less than half the increase in higher-rent units and far below the growing number of low-income renters for which these low-cost units would be affordable.

SEVERE GAPS IN SUPPLY With the private market failing to provide housing affordable to many of the nation’s lower-income households, the demand for low-cost rentals far outstrips supply. The shortfall is par-ticularly acute for extremely low-income renters (earning up to 30 percent of the area median), but also extends to very low-income renters (earning up to 50 percent of the area median). A National Low Income Housing Coalition study found that, in 2014, there were only 31 rental units affordable and available for every 100 extremely low-income renters, and 57 rental units affordable and available for every 100 very low-income renters (Figure 29).

Notes: Estimates include only units with cash rent reported. Total net change includes conversions to and from other uses, such

as seasonal and non-residential.

Source: JCHS tabulations of HUD, American Housing Surveys.

30

25

20

15

10

5

0

-5

-10Permanent

LossesFiltering

from OtherRent Levels

New Construction

Tenure Conversions

Total Net Change

Permanent Losses and the Slow Pace of Filtering Continue to Limit the Supply of Low-Rent UnitsComponents of Change in the Rental Stock, 2003–2013 (Percent)

FIGURE 28

Monthly Rent ● Under $800 ● $800 and Over


While large everywhere, the gap is especially wide in many fast-er-growing metros of the South and West. For example, Austin, Dallas, Las Vegas, Los Angeles, Orlando, Phoenix, Portland, Riverside, Sacramento, and San Diego all had no more than one affordable and available unit for every five extremely low-income renter households living in the area. The housing shortage for extremely low-income renters is most acute in the New York (610,000 units) and Los Angeles (382,000 units) metro areas.

Affordable rentals that can accommodate larger families are particularly difficult to find. As a result, the share of four-or-more-person renter families with children that were living in crowded conditions (more than two persons per bedroom) was nearly 19 percent in 2013, compared with an overall share of renter households of 5.5 percent. The incidence of overcrowd-ing among large families classified as very low income is even higher at 25 percent.

One obvious reason for overcrowding is that larger rental units are generally more expensive than smaller units. Even more important, though, many of the larger units afford-able to extremely low-income households are occupied by higher-income households. This includes 52 percent of three-bedroom units affordable to four-or-more-person households with extremely low incomes, 23 percent of two-bedroom units affordable to three-person households, and 28 percent of stu-dios or one-bedroom units affordable to two-person households.

Accessible rentals are also in short supply. As of 2011, less than 40 percent of the rental stock had no-step entries, and only 7 percent had extra-wide halls and doors allowing wheelchair access. In total, just 1 percent of rental units offered these fea-tures as well as single-floor living, lever-style door handles, and accessible electrical controls. And although newer rentals in larg-er multifamily buildings are somewhat more likely to include these five basic accessibility features, their pricing is often out of reach for low-income elderly and disabled households.

PERSISTENT MARKET TIGHTENING The inability of supply to keep up with the rapid rise in demand has led to the longest period of rental market tightening since the late 1960s. Starting in late 2010, the national rental vacancy rate fell for five consecutive years, hitting just 7.1 percent in 2015—its lowest point since 1985. In 2014–2015 alone, the vacancy rate for multifamily buildings with five or more units edged down another 0.9 percentage point while that for single-family rentals slipped 0.2 percentage point.

Growth in the Consumer Price Index for rent of primary resi-dence continued through 2015, far outstripping overall inflation (Figure 30). This is a marked departure from the long-run trend, with rent increases averaging slightly below general inflation since the 1960s. Nominal rents continued their climb through March 2016, with annual rates of increase pushing 3.7 percent.

20

15

10

5

0

AffordableUnits

AffordableUnits

Very Low-Income Renters

Extremely Low-Income Renters

FIGURE 29

● Unavailable ● Available

Low-Income Renters Far Outnumber the Supply of Available Units They Can AffordHouseholds and Units in 2014 (Millions)

Notes: Extremely/very low-income households earn no more than 30%/50% of area medians. Affordable units have rents up to 30% of household income after adjusting for household and unit sizes.Source: JCHS tabulations of National Low Income Housing Coalition, The Gap: The Affordable Housing Gap Analysis 2016.

Source: JCHS tabulations of US Bureau of Labor Statistics and MPF Research data.

6

5

4

3

2

1

0

-1

-2

-3

-4

-5

-6

2005 2006 2007 2008 2009 2010 2011 2012 2013

● Rent Index for Primary Residence ● Rents for Professionally Managed Apartments

● Prices for All Consumer Items

2014 2015

Rents Continue to Climb Despite Unusually Low Price InflationAnnual Change (Percent)

FIGURE 30


At the metro level, rent inflation ranged from under 2.0 percent in Cleveland, Philadelphia, and Washington, DC, to 6.0 percent or more in Houston, San Francisco, and Seattle.

The professionally managed apartment sector remains tight in most major markets, with MPF Research reporting a vacancy rate of just 4.2 percent in the first quarter of 2016—a 30 basis-point decline from a year earlier. Much of the recent tightening occurred within the two lower tiers (Class B and C) of the mar-ket. Overall vacancy rates varied widely from 3.0 percent or less in Los Angeles, Miami, Minneapolis, New York, Portland, and Sacramento, to more than 6.0 percent in Houston, Indianapolis, and Memphis. While more than two-thirds of the 94 metro areas that MPF Research tracks reported lower vacancies in the first quarter of 2016, a few major markets—such as Denver, Houston, and Pittsburgh—saw an uptick in rates from a year earlier.

Nationwide, rents in the professionally managed apartment sector rose by a strong 5.0 percent in the first quarter of 2016, up from 4.5 percent a year earlier. Increases were widespread, with rents in nearly all 94 metro markets on the rise. At the same time, however, rent growth slowed in a few areas, including Denver and Houston. In 18 of the nation’s 25 largest markets, rent increases in the middle tier (Class B) outstripped those in the upper tier (Class A).

STRONG MULTIFAMILY PERFORMANCEInvestor returns on rental properties continued to climb last year. The National Council of Real Estate Investment Fiduciaries reports that net operating income for commercial-grade apart-

ments increased for the fifth consecutive year in 2015, up nearly 11 percent from 2014. The annual rate of return on rental prop-erty investments rose to 12 percent, driven in large part by price appreciation. This strong performance has attracted investor demand, pushing capitalization rates for apartment properties down to 4.8 percent by year-end—the lowest level since the third quarter of 2008.

According to Moody’s/RCA Commercial Property Price Index, prices for apartment properties rose 13 percent in 2015, mark-ing the sixth consecutive year of double-digit growth. As of March 2016, apartment property prices stood 39 percent above their previous peak in late 2007. By comparison, the CoreLogic index indicated that single-family prices remained 5 percent below their pre-recession high. Prices for apartment properties in highly walkable central business districts increased the most last year (19 percent), while those in car-dependent suburbs rose somewhat more slowly (13 percent).

While strong nearly everywhere, apartment property prices in certain markets have skyrocketed. As of the fourth quarter of 2015, prices in the New York metro area stood 93 percent above their previous peak, while those in San Francisco were up 85 percent (Figure 31). Apartment prices in Boston, Denver, and Washington, DC, also topped previous peaks by more than 50 percent. In contrast, property prices in Las Vegas and Phoenix were up more modestly, likely because of the large oversupply of single-family homes available to meet rental demand.

With rental property prices on the rise, delinquency rates for most types of multifamily loans fell in 2015. The share of mul-tifamily loans held by FDIC-insured institutions that were at least 90 days past due or in non-accrual status dipped to just 0.28 percent in the fourth quarter, down from 0.44 percent a year earlier. In addition, the Mortgage Bankers Association indi-cates that 60-day delinquency rates for commercial/multifamily loans held by life insurance companies were at 0.04 percent, comparable to rates for loans held by Fannie Mae (0.07 percent) and Freddie Mac (0.02 percent).

Multifamily loans held in commercial mortgage-backed secu-rities (CMBS) posted a sharp drop in what had previously been relatively high delinquency rates. According to Moody’s Delinquency Tracker, the share of CMBS loans that were 60 or more days past due, in foreclosure, or in the lender’s possession peaked at nearly 16 percent in early 2011 before steadily retreat-ing to about half that share at the end of 2015. The share then fell to 2.1 percent in early 2016, but still more than double the 0.9 percent average in 2001–2007.

Unusually strong market conditions and historically low inter-est rates helped to propel a sharp rise in multifamily loan origi-nations last year. The MBA Originations Index indicates that the dollar volume of multifamily loans originated increased 31 percent in 2015. Meanwhile, total loans outstanding (including

Source: Moody’s Investors Service and Real Capital Analytics, Commercial Property Price Index for Apartments.

● New York ● San Francisco ● US ● Phoenix ● Las Vegas

550

500

450

400

350

300

250

200

150

100

50

0

2003 2005 2007 2009 2011 2013 20152001

Apartment Property Prices in Hot Markets Have Surged Well Above Previous PeaksApartment Price Index

FIGURE 31


both originations and repayments/write-offs) shot up by nearly $100 billion, to more than $1 trillion.

Bank and thrift balances rose by $47 billion (16 percent) in nomi-nal terms over the past year, while debt backed by federal sourc-es increased by $48 billion (11 percent). The federal government held or guaranteed 45 percent of all outstanding multifamily mortgage debt in 2015, a large share by historical standards. With Fannie Mae and Freddie Mac still in conservatorship after nearly eight years, the government’s future footprint in the multifamily lending market remains an open question.

THE OUTLOOKRental demand is expected to remain robust over the next decade as the youngest members of the millennial genera-tion reach their 20s and begin to form their own households. Moreover, if homeownership rates for households in their 30s and 40s continue to slide, rental demand will be stronger still. For their part, the aging baby-boom generation will boost the

number of older renters, ultimately pushing up demand for accessible units.

It is unknown whether high-income households will continue to fill the growing inventory of higher-end rentals or make the transition to homeownership. Regardless, expanding the rental supply through new market-rate construction should provide some slack to tight markets as older units slowly filter down from higher to lower rents. Once high-end demand is sated, developers in some areas may turn their attention to middle-market rentals, although high development costs mean that building new units affordable to even moderate-income house-holds is difficult without government subsidies.

And without public subsidies, the cost of a typical market-rate rental unit will remain out of reach for the nation’s lowest-income households. Indeed, with housing assistance insufficient to help most of those in need, the limited supply of low-cost units promises to keep the pressure on all renters at the lower end of the income scale.

HOUSING CHALLENGES


PREVALENCE OF COST BURDENS After three consecutive years of declines, the total number of housing cost-burdened households (paying more than 30 per-cent of income for housing) ticked up to 39.8 million in 2014. More than a third of US households faced cost burdens, includ-ing 16.5 percent with severe burdens (paying more than 50 per-cent of income for housing).

Driving this increase is the growing number of cost-burdened renters, which jumped from 20.8 million in 2013 to a record 21.3 million in 2014 (Figure 32). Worse still, more than half of these renters—11.4 million households—were severely burdened. These affordability pressures reflect the divergence between renter housing costs and renter incomes since 2001, with real median rental costs climbing 7 percent and real median renter incomes falling 9 percent.

Meanwhile, the number of cost-burdened homeowners declined for the fourth straight year in 2014, down 2 percent. This brought the share of cost-burdened homeowners to 25 percent, its lowest point in over a decade. Unlike renter hous-ing costs, owner housing costs fell 13 percent between 2010 and 2014, thanks in part to low interest rates but also to the fact that foreclosures forced many cost-burdened owners out of their homes.

Cost burdens remain nearly universal among lowest-income households (earning under $15,000), with 83 percent paying more than 30 percent of their incomes for housing in 2014. Most of these households were severely burdened, including 72 per-cent of renters and 66 percent of owners.

But households with moderate incomes are also burdened by high housing costs. Indeed, the cost-burdened rate among rent-ers earning $30,000–44,999 edged up from 47 percent in 2010 to 48 percent in 2014, while the cost-burdened rate among renters earning $45,000–74,999 held at the 2010 peak of 21 percent. Moreover, in the 10 metros with the highest median housing costs, three-quarters of renter households earning $30,000–44,999 and half of those earning $45,000–74,999 were cost burdened in 2014.

While easing among home-

owners, housing cost burdens are

a fact of life for a growing number

of renters. These burdens put

households at risk of housing

instability and homelessness,

particularly in the nation’s high-

cost cities. Meanwhile, growing

income inequality and the

concentration of poverty have

fueled an increase in residential

segregation. With dwindling

federal subsidies, state and local

governments are struggling

to preserve and expand the

supply of good-quality affordable

housing in all neighborhoods.

Reducing carbon emissions from

the residential sector is not only

a national challenge but a global

imperative.


CONSEQUENCES OF HIGH HOUSING COSTSAfter paying large shares of their incomes for housing, cost-burdened households cut back spending on other vital needs. According to the 2014 Consumer Expenditure Survey, severely burdened households in the bottom expenditure quartile (a proxy for low income) had just $500 left over to cover all other monthly expenses, while otherwise similar households living in affordable housing had more than twice that amount to spend. As a result, severely cost-burdened households spent 41 percent less on food and 74 percent less on healthcare than their coun-terparts living in housing they could afford.

To avoid cost burdens, low-income households often trade off location for affordability. In consequence, low-income house-holds living in housing they can afford spend nearly three times more on transportation than households with severe burdens. Low-income households without cost burdens are also more likely to live in inadequate units (Figure 33).

Very low-income renters (earning up to 50 percent of area medi-an) with severe burdens are at high risk of housing instability. In 2013, 11 percent of these households reported they had missed at least one rent payment within the previous three months, and 18 percent had either received a shutoff notice or had their utilities shut off for nonpayment. Furthermore, 9 percent stated that they were likely to be evicted within the next two months. Very low-income owners with severe burdens also faced these hardships, with 11 percent missing at least one mortgage pay-

ment within the previous three months and 10 percent having received a shutoff notice or had their utilities shut off.

One possible outcome for these vulnerable households is home-lessness, particularly if they live in the nation’s high-cost coast-al cities. Although overall homelessness fell 11 percent between 2010 and 2015, to about 565,000 people, the problem in some cities has reached crisis proportions. Indeed, more than one in five homeless people live in New York City or Los Angeles. In 2014–2015 alone, the homeless population in New York City increased by 11 percent and in Los Angeles by 20 percent.

Progress in eliminating homelessness varies widely across vulnerable populations. Thanks to targeted federal funding, homelessness among veterans fell by 36 percent between 2010 and 2015, and several cities—including Houston, New Orleans, and Philadelphia—have even declared an end to homelessness among this group. Chronic homelessness also fell 22 percent in 2010–2015, due largely to the expansion of permanent sup-portive housing, which offers services to address the mental health and substance abuse issues common to this population. The reduction in homelessness among people in families with children, however, has been much smaller (Figure 34).

One possible solution to family homelessness is to improve access to permanent housing subsidies. As HUD’s Family Options study has demonstrated, families leaving homeless shelters with housing vouchers are more than twice as likely as

Notes: Moderately/severely cost-burdened households pay more than 31–50% of income for housing. Households with zero or negative income are assumed to be severely burdened, while renters paying no cash rent are assumed to be without burdens.Source: JCHS tabulations of US Census Bureau, American Community Survey 1-Year Estimates.

Owners Renters

● Moderately Burdened ● Moderately Burdened

● Severely Burdened ● Severely Burdened

25

20

15

10

5

0

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

While the Number of Cost-Burdened Owners Has Fallen, the Number of Cost-Burdened Renters Has Reached a New HighHouseholds (Millions)

FIGURE 32


those without vouchers to remain stably housed. Accordingly, President Obama has proposed $11 billion in mandatory funding in his FY2017 budget for a new 10-year initiative to end home-lessness among families with children, significantly expanding housing choice vouchers and rapid rehousing assistance.

SHORTFALLS IN THE AFFORDABLE SUPPLYBetween 1993 and 2013, the number of very low-income house-holds eligible for federal rental housing assistance soared by 3.8 million, bringing the total to 18.5 million. Over this same period, however, the number of assisted renters rose by just 532,000 (Figure 35). As a result, the share of income-qualified renters that received assistance dropped from 29 percent to 26 percent.

With demand far outstripping supply, competition for housing assistance is intense. The waiting lists for housing vouchers managed by local public housing authorities (PHAs) are years long or even closed. According to HUD’s Picture of Subsidized Households, a renter household that used a voucher in 2015 had waited more than two years on average to move into a unit, with the wait time in the San Diego metro area as long as seven years.

The US Treasury Department’s Low Income Housing Tax Credit (LIHTC) program, the primary vehicle for expanding the afford-able housing supply, has supported construction and preserva-tion of roughly 2.8 million rental units since 1986. The tax cred-its are allocated to states on a per capita basis, and applications

Note: The chronically homeless have been without a place to live for at least a year or have had repeated episodes of

homelessness over the past few years.

Source: JCHS tabulations of HUD, 2015 Annual Homeless Assessment Report to Congress.

30

20

10

0

-10

-20

-30

-40

People in Families

with Children

Chronically Homeless

Individuals

Veterans

● 2007–2010 ● 2010–2015

Progress in Reducing Family Homelessness Has Been Comparatively Modest Change in Homeless Population (Percent)

FIGURE 34

Notes: Extremely low/very low/low income is defined as up to 30%/31–50%/51–80% of area medians. Cost-burdened households pay more than 30% of income for housing costs. Inadequate units lack complete bathrooms, running water, electricity, or have other serious deficiencies. Source: JCHS tabulations of HUD, 2013 American Housing Survey.

● Not Burdened ● Cost Burdened

14

12

10

8

6

4

2

0

14

12

10

8

6

4

2

0

Extremely Low Very Low Low All Higher Extremely Low Very Low Low All Higher

Income LevelIncome Level

Many Low-Income Households Sacrifice Housing Quality for AffordabilityShare of Renters Living in Inadequate Units (Percent)

FIGURE 33

Share of Owners Living in Inadequate Units (Percent)


for the credits far exceed available funding. By itself, though, the tax credit is insufficient to support development of hous-ing affordable to the nation’s lowest-income households and is therefore often combined with other subsidies, like those under the HOME program. The 55 percent real reduction in HOME funding between FY2006 and FY2016 has thus eroded the power of the LIHTC program to add new affordable rentals.

Faced with shrinking federal resources, state and local govern-ments are attempting to fill the financing gaps. A report by the Technical Assistance Collaborative found that 30 states offered some form of state-funded rental assistance in 2014, with annu-al funding ranging from about $5 million in Delaware to $83 million in Massachusetts. At the local level, cities have turned to a variety of alternative financing methods, such as taxes on real estate transactions, tax-increment financing, and linkage fees on commercial development.

Cities have also adopted or revised their inclusionary housing ordinances, either mandating that a share of units in new hous-ing developments over a certain size be affordable to low- and moderate-income households or offering density bonuses in exchange for setting aside affordable units. According to a 2014 report by the Lincoln Institute of Land Policy, inclusionary hous-ing programs exist in more than 500 local jurisdictions. These programs typically provide long-term affordability, which is important in high-cost areas and in gentrifying neighborhoods where low-income households are at risk of displacement.

But local land use regulations—such as zoning requirements, density and height restrictions, and minimum lot size and park-

ing requirements—can also inhibit construction of affordable housing in expensive metro areas. For example, a 2008 study by Harvard’s Rappaport Institute for Greater Boston found that a one-acre increase in a local town’s minimum lot size was asso-ciated with about a 40 percent drop in housing permits.

High land and wage costs also deter affordable housing devel-opment. A 2015 Urban Land Institute report estimated that in hot housing markets, land costs for a high-rise, mixed-income project with affordable units could account for as much as 25 percent of total development costs. Similarly, the Citizens Housing and Planning Council in New York City estimated that a prevailing wage requirement for affordable housing projects in 2011 could also raise development costs by roughly 25 per-cent. These added costs must be met with either an increase in government subsidies or a reduction in affordable units.

These conditions make preservation of the existing supply of assisted housing all the more urgent. According to the National Housing Preservation Database, the affordable-use restrictions on nearly 2 million federally assisted rental units will expire over the coming decade. A majority (64 percent) of this at-risk stock is supported through the LIHTC program, which is approaching its 30-year anniversary.

On the public housing side, the Rental Assistance Demonstration (RAD) has given PHAs new flexibility to use tax credits and pri-vate capital to rehabilitate and preserve the aging inventory. As of December 2015, HUD estimates that PHAs and their part-ners raised over $1.7 billion through RAD to convert more than 26,000 public housing units to long-term contracts.

Note: Very low income is defined as 50% or less of area median.Source: JCHS tabulations of HUD, Worst Case Housing Needs Reports to Congress.

● Unassisted ● Assisted

20.0

17.5

15.0

12.5

10.0

7.5

5.0

2.5

0

1983 19891987 1993 1995 19971991 1999 2001 20031985 20072005 20112009 2013

After Some Increase in the 1980s, the Number of Assisted Households Has Been Essentially Flat for Two Decades Very Low-Income Renter Households (Millions)

FIGURE 35


INCREASING POVERTY AND RESIDENTIAL SEGREGATION Poverty in the United States has become more concentrated. In 2014, 13.7 million people lived in neighborhoods with poverty rates of 40 percent or higher, up from 6.5 million in 2000. This jump largely reflects widespread declines in incomes since the start of the last decade as well as increasing residential segrega-tion by income.

The location of assisted housing in low-income communities has contributed to this pattern. Public housing is most likely to be located in high-poverty neighborhoods, a legacy of develop-ments built in the 1940s and 1950s in economically and racially segregated neighborhoods. Decades later, 35 percent of public housing units are in census tracts with at least a 40 percent poverty rate, and another 42 percent are in tracts with 20–39 percent rates. By comparison, just 15–18 percent of rentals sub-sidized through the housing voucher and LIHTC programs are located in high-poverty census tracts.

The Supreme Court’s ruling on disparate-impact claims in 2015 may help to limit further concentration of affordable housing in high-poverty areas. In addition, HUD issued a final rule on Affirmatively Furthering Fair Housing requiring state and local recipients of HUD funds, as well as all PHAs, to identify patterns of segregation and develop concrete steps to foster greater inte-gration. Even before last year, however, several states—includ-ing Massachusetts, New Jersey, and Pennsylvania—had made progress in lifting the share of LIHTC units built in low-poverty communities by giving higher priority to projects in these loca-tions in their tax credit allocations.

These efforts, however, must be viewed within the context of increasing residential segregation by income. Research from the Stanford Center for Education Policy Analysis shows that, from 1970 to 2012, the share of families living in middle-income neighborhoods plummeted by 24 percentage points while the shares living in low- and high-income neighborhoods increased by 11 and 13 percentage points, respectively (Figure 36). Growing socioeconomic segregation has significant negative conse-quences for the families left in neighborhoods with limited pub-lic services and unsafe living conditions.

Exclusionary zoning in the form of density restrictions and complex municipal review processes help to reinforce the isolation of low-income households. While states and locali-ties have enacted legislation to eliminate exclusionary zoning practices and encourage inclusionary development, the scale of these efforts falls far short of the millions of affordable units produced through the LIHTC and HOME programs. Indeed, the Lincoln Institute of Land Policy estimates that inclusionary housing programs had produced just 129,000–150,000 afford-able units nationwide as of 2010.

HOUSING NEEDS IN RURAL AREAS AND NATIVE LANDS Households living outside metropolitan areas have their own set of housing challenges. Poverty is widespread, affecting 18 percent of the non-metro population and 29 percent of people living in tribal areas. Indeed, poverty rates across all age and racial/ethnic groups are higher in non-metro than metro areas. In 2013, 41 percent of very low-income homeowners in non-metro areas were severely housing cost burdened, along with 48 percent of very low-income renters.

Substandard housing is a particular problem in these areas. Compared with the typical US unit, housing in non-metro areas is two times more likely to have incomplete plumbing, while housing in tribal tracts is five times more likely to lack this basic function (Figure 37). While manufactured homes can be an important source of affordable housing in non-metro areas, 29 percent of the occupied stock in 2013 was built before HUD set federal design and construction standards in 1976. Of these older homes, 8 percent are categorized as inadequate.

Yet another housing challenge in rural areas is the high concen-tration of older adults, with 17 percent of the non-metro popu-lation age 65 and over compared with 13 percent of the metro population. The supply of accessible housing in these areas is limited, with just under a third having both no-step entries and single-floor living.

Meanwhile, federal housing assistance for non-metro house-holds remains modest. As of 2012, USDA halted new construc-tion of affordable rental housing through Section 515, leaving only the Section 514/516 Farmworker Housing program to finance new units in rural areas. In addition, using the LIHTC

Notes: Low-/middle-/high-income census tracts have median incomes under 80%/80–125%/more than 125% of metro area medians. Data include 117 metro areas with populations of 500,000 or more in 2007.Source: K. Bischoff and S. Reardon, The Continuing Increase in Income Segregation, 2007–2012, Stanford Center for Education Policy Analysis, 2016.

Census Tract Type ● Low Income ● Middle Income ● High Income

1970 1980 1990 2000 2012

70

60

50

40

30

20

10

0

Income Segregation Has Steadily Increased Over the Last Four DecadesShare of Family Households (Percent)

FIGURE 36


program to expand the supply of affordable rental housing in these areas can be difficult, given that states generally give pri-ority to projects located near public transit and services. Federal assistance for rural homeowners is also increasingly limited, with funding for USDA’s Section 502 direct loan program falling from $34 million in FY2005 to about $28 million in FY2015.

RESIDENTIAL CARBON EMISSIONS AND ENERGY USEWith the signing of the Paris Climate Agreement in December 2015, President Obama committed to reducing US greenhouse gas emissions to 2005 levels by 2025. To meet this goal, policy-makers must prioritize large cutbacks in the residential sector, which accounts for over a fifth of national carbon emissions.

The largest reductions in energy use can be achieved by retrofit-ting the existing stock. While the upfront investment required may be an obstacle for some property owners, tax credit and rebate programs can promote upgrades. Indeed, 63 percent of respondents to the 2015 Demand Institute Consumer Housing Survey stated that incentives were important to their likelihood of making energy-efficient improvements.

To encourage rental property owners to retrofit their units, FHA recently reduced its insurance rates on mortgages for multi-family properties meeting federal green building and energy performance standards. In addition, a number of state housing finance agencies currently provide loans for efficiency upgrades to both single-family and multifamily housing.

These efficiency improvements can yield important savings for low-income households, who pay much larger portions of their incomes for utilities than high-income households. For example, renter households earning under $15,000 a year in 2014 devoted 17 percent of their incomes to utility payments, and owner households with similar incomes paid 22 percent. By comparison, utility costs for both owners and renters earning at least $75,000 a year amounted to just 2 percent of income.

Meanwhile, development patterns play a large role in transporta-tion emissions, which are responsible for 34 percent of total emis-sions. According to a 2014 University of California Berkeley study, suburban households have a larger carbon footprint than urban or rural households not only because of their larger homes but also because of their higher rates of vehicle ownership. Similarly, a 2015 Boston University analysis found that lower-density met-ros like Denver and Salt Lake City have higher carbon emissions per capita than older, higher-density cities.

State and local efforts may be instructive to federal policymak-ers. Changes in the International Energy Conservation Code have already led to tighter state and local standards for new construction and remodeling. For its part, California has taken a leading role in reducing greenhouse gases by adopting the Clean Energy and Pollution Reduction Act of 2015, requiring a 50 percent increase in the energy efficiency of existing buildings by 2030.

THE OUTLOOKIn 2016, after an eight-year delay, HUD allocated nearly $174 million to states through the National Housing Trust Fund—the first new program to expand the supply of affordable hous-ing for extremely low-income renters in a generation. While these funds will give a much-needed boost to state and local programs, the growing gap between the rents for new units and the amounts lowest-income households can afford to pay for housing underscores the difficulty of increasing the afford-able supply through new construction alone. Current proposals to expand the LIHTC program, as well as to reform the public housing and other rental assistance programs, may help broad-en access to affordable housing for the nation’s most vulnerable households. But preserving and maintaining the private supply of low-cost housing—where the majority of low-income renters live—is also crucial.

Reducing residential segregation by income will involve a con-certed effort by federal, state, and local governments to foster more equitable access to opportunity for people of all races and incomes. While reducing the growing isolation of the poor is key, addressing the self-segregation of the wealthy is also essential. At the same time, however, new investments in low-income communities—including job training, school quality, and healthcare facilities and other services—are no less critical to the well-being of millions of families.

Notes: Tribal census tracts are as defined by the US Census Bureau for 2010. Rural census tracts are in non-metro areas. Source: JCHS tabulations of US Census Bureau, 2010–2014 American Community Survey 5-Year Estimates.

Population in Poverty Units Lacking Complete Plumbing

Units Lacking Complete Kitchens

30

25

20

15

10

5

0

Census Tract Type ● Tribal ● Rural ● All

Residents of Rural Areas and Tribal Lands Are Especially Likely to Live in Poverty and Have Substandard HousingShare of Population and Housing Units (Percent)

FIGURE 37

APPENDIX TABLES


Table A-1 ........... Housing Market Indicators: 1980–2015

Table A-2 ........... Housing Cost-Burdened Households by Tenure and Income: 2001, 2013, and 2014

Additional tables can be downloaded in Microsoft Excel format at www.jchs.harvard.edu, including:

Table W-1 .......... Homeownership Rates by Age, Race/Ethnicity, and Region: 1994–2015

Table W-2 .......... Housing Cost-Burdened Households by Demographic Characteristics: 2014

Table W-3 .......... Monthly Housing and Non-Housing Expenditures by Households: 2014

Table W-4 .......... Metro Area Monthly Mortgage Payment on the Median Priced Home: 1990–2015

Table W-5 .......... Metro Area Cost Burden Rates by Household Income: 2014


Housing Market Indicators: 1980–2015

TABLE A-1

Notes: All value series are adjusted to 2015 dollars by the CPI-U for All Items. All links are as of May 2016. n/a indicates data not available. Sources:1. US Census Bureau, New Privately Owned Housing Units Authorized by Building Permits in Permit-Issuing Places, http://www.census.gov/construction/nrc/xls/permits_cust.xls. 2. US Census Bureau, New Privately Owned Housing Units Started, https://www.census.gov/construction/nrc/xls/starts_cust.xls. 3. US Census Bureau, Shipments of New Manufactured Homes, http://www.census.gov/construction/mhs/pdf/shiphist.pdf & http://www.census.gov/construction/mhs/xls/shipmentstostate11-15.xls. Data from 1980-2010 retrieved from JCHS historical tables.

Manufactured housing starts are defined as shipments of new manufactured homes. 4. US Census Bureau, New Privately Owned Housing Units Completed in the United States by Purpose and Design, http://www.census.gov/construction/nrc/xls/quarterly_starts_completions_cust.xls and JCHS historical tables.5. New home price is the median price from US Census Bureau, Median and Average Sales Price of New One-Family Houses Sold, http://www.census.gov/construction/nrs/xls/usprice_cust.xls.

Year

Permits 1

(Thousands)

Starts

(Thousands)

Size 4

(Median sq. ft.)

Median Sales Price of

Single-Family Homes

(2015 dollars)

Single-Family Multifamily Single-Family 2

Multifamily 2

Manufactured 3

Single-Family Multifamily New 5

Existing 6

1980 710 480 852 440 222 1,595 915 185,817 178,482

1981 564 421 705 379 241 1,550 930 179,653 172,417

1982 546 454 663 400 240 1,520 925 170,210 166,280

1983 901 704 1,068 636 296 1,565 893 179,191 166,162

1984 922 759 1,084 665 295 1,605 871 182,268 164,931

1985 957 777 1,072 670 284 1,605 882 185,693 165,996

1986 1,078 692 1,179 626 244 1,660 876 198,956 173,564

1987 1,024 510 1,146 474 233 1,755 920 218,031 178,545

1988 994 462 1,081 407 218 1,810 940 225,397 178,714

1989 932 407 1,003 373 198 1,850 940 229,371 180,264

1990 794 317 895 298 188 1,905 955 222,872 175,602

1991 754 195 840 174 171 1,890 980 208,826 177,545

1992 911 184 1,030 170 211 1,920 985 205,257 177,467

1993 987 213 1,126 162 254 1,945 1,005 207,492 177,558

1994 1,068 303 1,198 259 304 1,940 1,015 207,910 180,282

1995 997 335 1,076 278 340 1,920 1,040 208,245 180,147

1996 1,069 356 1,161 316 363 1,950 1,030 211,487 184,120

1997 1,062 379 1,134 340 354 1,975 1,050 215,604 189,084

1998 1,188 425 1,271 346 373 2,000 1,020 221,749 196,254

1999 1,247 417 1,302 339 348 2,028 1,041 229,050 199,530

2000 1,198 394 1,231 338 250 2,057 1,039 232,613 200,967

2001 1,236 401 1,273 329 193 2,103 1,104 234,474 206,782

2002 1,333 415 1,359 346 169 2,114 1,070 247,162 218,956

2003 1,461 428 1,499 349 131 2,137 1,092 251,186 229,696

2004 1,613 457 1,611 345 131 2,140 1,105 277,294 241,921

2005 1,682 473 1,716 353 147 2,227 1,143 292,357 263,949

2006 1,378 461 1,465 336 117 2,248 1,172 289,805 260,864

2007 980 419 1,046 309 96 2,277 1,197 283,380 246,362

2008 576 330 622 284 82 2,215 1,122 255,508 215,520

2009 441 142 445 109 50 2,135 1,113 239,407 190,566

2010 447 157 471 116 50 2,169 1,110 241,087 187,762

2011 418 206 431 178 52 2,233 1,124 239,399 173,789

2012 519 311 535 245 55 2,306 1,098 253,128 181,467

2013 621 370 618 307 60 2,384 1,059 273,586 200,840

2014 640 412 648 355 64 2,453 1,073 283,136 209,148

2015 696 487 715 397 71 2,467 1,074 296,400 223,900


6. Existing home price is the median sales price of existing single-family homes determined by the National Association of Realtors® (NAR), obtained from NAR and Moody’s Economy.com.7. US Census Bureau, Housing Vacancy Survey, http://www.census.gov/housing/hvs/data/ann15ind.html.8. US Census Bureau, Annual Value of Private Construction Put in Place, http://www.census.gov/construction/c30/historical_data.html; data 1980-1993 retrieved from past JCHS reports. Single-family and multifamily are new

construction. Owner improvements do not include expenditures on rental, seasonal, and vacant properties.9. US Census Bureau, Houses Sold by Region, http://www.census.gov/construction/nrs/xls/sold_cust.xls.10. National Association of Realtors®, Existing Single-Family Home Sales obtained from and annualized by Moody’s Economy.com, and JCHS historical tables.

Vacancy Rates 7

(Percent)

Value Put in Place 8

(Millions of 2015 dollars)

Home Sales

(Thousands)

For Sale For Rent Single-Family Multifamily Owner Improvements New 9 Existing 10

1.4 5.4 152,223 48,059 n/a 545 2,973

1.4 5.0 135,496 45,526 n/a 436 2,419

1.5 5.3 101,836 38,163 n/a 412 1,990

1.5 5.7 172,561 53,417 n/a 623 2,697

1.7 5.9 197,085 64,378 n/a 639 2,829

1.7 6.5 192,411 62,865 n/a 688 3,134

1.6 7.3 225,190 67,122 n/a 750 3,474

1.7 7.7 244,561 53,103 n/a 671 3,436

1.6 7.7 240,609 44,675 n/a 676 3,513

1.8 7.4 231,147 42,632 n/a 650 3,010

1.7 7.2 204,712 34,909 n/a 534 2,917

1.7 7.4 173,024 26,361 n/a 509 2,886

1.5 7.4 206,061 22,120 n/a 610 3,155

1.4 7.3 229,838 17,695 93,936 666 3,429

1.5 7.4 259,582 22,520 103,384 670 3,542

1.5 7.6 238,751 27,822 88,208 667 3,523

1.6 7.8 258,000 30,702 100,277 757 3,795

1.6 7.7 258,694 33,792 98,401 804 3,963

1.7 7.9 289,959 35,733 105,218 886 4,496

1.7 8.1 318,446 39,030 106,744 880 4,650

1.6 8.0 325,916 38,896 111,614 877 4,602

1.8 8.4 333,358 40,558 113,788 908 4,732

1.7 8.9 350,307 43,414 128,923 973 4,974

1.8 9.8 400,063 45,234 129,257 1,086 5,444

1.7 10.2 473,729 50,119 144,794 1,203 5,958

1.9 9.8 526,110 57,400 174,476 1,283 6,180

2.4 9.7 489,079 62,079 163,409 1,051 5,677

2.7 9.7 348,862 55,966 153,982 776 4,398

2.8 10.0 204,512 48,810 142,074 485 3,665

2.6 10.6 116,374 31,528 125,561 375 3,870

2.6 10.2 122,357 15,963 124,761 323 3,708

2.5 9.5 113,987 15,844 127,399 306 3,786

2.0 8.7 136,283 23,238 118,986 368 4,128

2.0 8.3 173,744 32,049 123,224 429 4,484

1.9 7.6 193,830 41,856 134,799 437 4,344

1.8 7.1 218,523 51,899 147,838 501 4,646


Housing Cost-Burdened Households by Tenure and Income: 2001, 2013, and 2014Thousands

TABLE A-2

Tenure and Income

2001 2013 2014

Not

Burdened

Moderately

Burdened

Severely

Burdened Total

Not

Burdened

Moderately

Burdened

Severely

Burdened Total

Not

Burdened

Moderately

Burdened

Severely

Burdened Total

Owners

Under $15,000 1,008 855 2,683 4,547 921 882 3,438 5,240 871 873 3,395 5,140

$15,000–29,999 4,314 1,803 1,816 7,933 4,325 2,220 2,320 8,865 4,160 2,227 2,296 8,683

$30,000–44,999 5,711 2,037 991 8,739 6,019 2,392 1,198 9,609 5,957 2,369 1,151 9,477

$45,000–74,999 13,100 3,293 723 17,116 13,179 3,084 816 17,079 13,216 3,015 764 16,996

$75,000 and Over 29,098 2,282 272 31,651 30,612 2,219 310 33,141 31,398 2,109 281 33,787

Total 53,231 10,270 6,485 69,986 55,055 10,797 8,082 73,933 55,602 10,593 7,888 74,083

Renters

Under $15,000 1,460 933 4,921 7,314 1,613 1,096 6,973 9,682 1,577 1,109 6,943 9,629

$15,000–29,999 2,369 3,218 2,101 7,688 2,283 3,921 3,352 9,555 2,189 3,851 3,517 9,557

$30,000–44,999 4,134 2,098 321 6,553 3,958 2,680 683 7,321 3,821 2,859 732 7,412

$45,000–74,999 7,390 900 102 8,392 6,754 1,504 197 8,455 6,880 1,652 211 8,743

$75,000 and Over 6,304 186 12 6,502 6,986 349 10 7,345 7,437 383 16 7,835

Total 21,658 7,335 7,457 36,450 21,593 9,549 11,216 42,358 21,905 9,854 11,418 43,176

All Households

Under $15,000 2,469 1,788 7,604 11,861 2,533 1,977 10,411 14,921 2,448 1,982 10,339 14,769

$15,000–29,999 6,683 5,021 3,918 15,622 6,608 6,141 5,672 18,421 6,350 6,078 5,812 18,241

$30,000–44,999 9,846 4,135 1,311 15,292 9,977 5,072 1,881 16,930 9,778 5,227 1,883 16,889

$45,000–74,999 20,490 4,193 825 25,508 19,933 4,587 1,013 25,534 20,096 4,668 975 25,739

$75,000 and Over 35,402 2,468 284 38,153 37,597 2,568 320 40,485 38,834 2,491 297 41,622

Total 74,889 17,605 13,942 106,436 76,648 20,345 19,297 116,291 77,507 20,447 19,306 117,259

Notes: Moderate (severe) burdens are defined as housing costs of more than 30% and up to 50% (more than 50%) of household income. Households with zero or negative income are assumed to be severely burdened, while renters paying no cash rent are assumed to be unburdened. Income cutoffs are adjusted to 2014 dollars by the CPI-U for All Items.Source: JCHS tabulations of US Census Bureau, American Community Surveys.

For additional copies, please contact

Joint Center for Housing Studies of Harvard University

One Bow Street, Suite 400

Cambridge, MA 02138

www.jchs.harvard.edu

twitter: @Harvard_JCHS

The Joint Center for Housing Studies of Harvard University advances

understanding of housing issues and informs policy. Through its research,

education, and public outreach programs, the center helps leaders in

government, business, and the civic sectors make decisions that effectively

address the needs of cities and communities. Through graduate and executive

courses, as well as fellowships and internship opportunities, the Joint Center

also trains and inspires the next generation of housing leaders.

STAFF

Kermit Baker

Pamela Baldwin

Heidi Carrell

James Chaknis

Matthew Curtin

Angela Flynn

Christopher Herbert

Jessica Jean-Francois

Elizabeth La Jeunesse

Mary Lancaster

Irene Lew

Ellen Marya

Sonali Mathur

Daniel McCue

Jennifer Molinsky

Shannon Rieger

Jonathan Spader

Alexander von Hoffman

Abbe Will

Georgia Williams

FELLOWS

Barbara Alexander

William Apgar

Michael Berman

Rachel Bratt

Michael Carliner

Kent Colton

Daniel Fulton

Aaron Gornstein

George Masnick

Shekar Narasimhan

Nicolas Retsinas

Mark Richardson

EDITOR

Marcia Fernald

DESIGNER

John Skurchak

Joint Center for Housing Studies of Harvard University

FIVE DECADES OF HOUSING RESEARCH SINCE 1959

State of the Nation’s Housing - Joint Center for Housing · PDF fileThe opinions expressed in The State of the Nation’s Housing 2016 do not necessarily represent the views of Harvard

Documents